the reference point effect, m&a misvaluations and merger...

Durham E-Theses

The Reference Point E�ect, M&A Misvaluations and

Merger Decisions

LI, ZHENLONG

How to cite:

LI, ZHENLONG (2017) The Reference Point E�ect, M&A Misvaluations and Merger Decisions , Durhamtheses, Durham University. Available at Durham E-Theses Online: http://etheses.dur.ac.uk/12225/

Use policy

The full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission orcharge, for personal research or study, educational, or not-for-pro�t purposes provided that:

• a full bibliographic reference is made to the original source

• a link is made to the metadata record in Durham E-Theses

• the full-text is not changed in any way

The full-text must not be sold in any format or medium without the formal permission of the copyright holders.

Please consult the full Durham E-Theses policy for further details.

Academic Support O�ce, Durham University, University O�ce, Old Elvet, Durham DH1 3HPe-mail: [email protected] Tel: +44 0191 334 6107

http://etheses.dur.ac.uk

http://www.dur.ac.uk

http://etheses.dur.ac.uk/12225/

http://etheses.dur.ac.uk/12225/

http://etheses.dur.ac.uk/policies/

http://etheses.dur.ac.uk

The Reference Point Effect,

M&A Misvaluations and Merger

Decisions

Zhenlong Li

Principal Supervisor: Dr. Jie (Michael) Guo

A Thesis Presented for the Degree of Doctor of

Philosophy in Finance

Durham University Business School

Durham University

Mar 2017

I

The Reference Point Effect,

M&A Misvaluations and Merger Decisions

Abstract

This thesis investigates the reference point effect on M&As. Prospect theory proposes that the

reference dependence bias is firmly rooted in people’s minds, affirming that people rely heavily

on a piece of relevant information while making decisions. This thesis applies this reference

point to M&As, showing that M&A participants are subject to reference-dependence bias. The

reference point theory explains the M&A motives in a new way. The results of this thesis offer

important implications to M&A practitioners.

Baker et al. (2012) suggest that the target’s reference price enhances the target’s bargaining

power in the U.S. market, inspiring the author’s thinking that the reference point effect is likely

to be reinforced in a competitive market and in the scenario where the bidder is in the face of

considerable information barriers. In this pursuit, Chapter 3 studies the reference point effect

on a sample of public acquisitions involving a U.K. target. Using the target 52-week high as

the reference price, a positive relationship emerges between the reference price and the offer

premium. Further, there is evidence of overpayment among domestic bidders whereas little

evidence among cross-border bidders, indicating that the market acquiesces the payment

according to the target reference point among cross-border bidders. Evidence that reference-

dependence bias serves as the bargaining power is confirmed outside the United States.

However, the real M&A motive of the bidder is unlikely to be revealed solely depending on

the target reference point. Therefore, Chapter 4 extends Baker et al.’s paper (2012) by adding

both the target and the bidder reference points. A Relative Reference Point (RRP) is proposed

for M&A misvaluations as per Shleifer and Vishny (2003), indicating the extent to which the

bidder is more overvalued relatively to the target. The joint reference point effects can explain

the motivation of why bidders paying high offer premiums. The results obtained in this chapter

show that bidders are likely to pay with stocks when RRP increases, suggesting that bidders

accelerating the process of overvaluation dilution is a sign of bidder overvaluation. In addition,

the offer premium increases with the RRP, indicating the motive of diluting overvaluation is to

be seen by the target, leading targets to bargain over high offer premiums. When assessing the

long-term performance of stock bidders, it became evident that bidders time the market by

paying stocks according to the RRP.

A direct implication of prospect theory is that people seek risk in the face of loss gauged by a

reference point. Chapter 5 proposes the bidder reference point to assess the market anticipation

effect. Investors suffer mental loss when their firm’s current performance is significantly below

its best recently achieved performance. They anticipate that the firm will take risks to turn the

table round. It was evident that the firm is rewarded with positive market reactions when it

takes risks according to the market’s anticipation. In a further analysis, it also emerged that

managers taking the market-anticipated risks exhibit greater efforts in the processes of M&A

negotiation and post-merger integration, reflected in the low offer premiums paid to the target

shareholders and positive long-term abnormal returns earned.

This thesis has empirically investigated many implications of the reference point effect in the

M&A context. The main aim of this thesis is to simplify M&A decisions for managers in order

to structure M&A effectively.

II

No part of this thesis has been previously submitted for a degree in this and other university.

The copyright of this thesis rests with the author or the University to which it was submitted. No quotation from

it, or information derived from it should be published without the written consent of the author or the University,

and information derived from it should be acknowledged.

III

To my parents

IV

Acknowledgements

There are many people have helped me considerably in the completion of my Ph.D studies.

First, I would like to give the greatest gratitude and appreciation to my principal supervisor Dr.

Jie (Michael) Guo who brought me into academia. He offered me valuable advice and

knowledge throughout my Ph.D journey. He also encouraged me to attend high quality

academic conferences and recommended me to work in the financial institution. Those

experiences are of great benefits to my professional career and research development. My Ph.D

journey would not go thus far if without his encouragement and generous support. I would like

to extend my thanks to Michael. It is my greatest fortune to be part of his Ph.D team.

I am also grateful to Dr. Panagiotis Andrikopoulos, the co-author for one of my working papers,

for his advice in improving the quality of my thesis, and I would also like to express my

gratitude to Dr. Evangelos Vagenas-Nanos who offered me many thoughts to improve the

fourth Chapter of this thesis.

A huge thank you to my deeply beloved parents, my father Li, Yong and my mother Zheng,

Defeng, whose love to me are unconditional. They funded my Ph.D studies and encouraged

me to move on with a brave and humble heart. There are no words in this world could describe

how much I love you both.

Special thanks to those who have helped me in this wonderful journey. Dr. Kai Lisa Lo, Dr

Yichen Li and Xiaofei Xing, who offered me great support in research. Rev. Lawrence Lo,

whose words of wisdom taught me to deal with the difficulties of life. And Dr. Eason Chan,

whose music inspired me. A number of people, whose names though not listed in the

acknowledgements, their cares to me are unforgettable and the motivation towards the

completion of my Ph.D studies.

Zhenlong Li

14th Feb, 2017

V

Table of Contents

ABSTRACT ............................................................................................................................... I

ACKNOWLEDGEMENTS ..................................................................................................... IV

CHAPTER 1: INTRODUCTION .............................................................................................. 1

CHAPTER 2: LITERATURE REVIEW ................................................................................. 16

2.1. M&A motives ................................................................................................................... 17

2.1.1. Synergies ........................................................................................................................ 18

2.1.2. Agency problems ............................................................................................................ 22

2.1.3. Behavioural finance theories ......................................................................................... 25

2.1.3.1. Managerial overconfidence ........................................................................................ 26

2.1.3.2. Investors’ overreactions and underreactions ............................................................. 30

2.1.3.3. The misvaluation hypothesis ....................................................................................... 32

2.2. M&A process .................................................................................................................... 38

2.2.1. Selecting a target ........................................................................................................... 38

2.2.2. Paying for a target ......................................................................................................... 41

2.2.2.1. M&A offer premiums .................................................................................................. 42

2.2.2.2. Method of payment hypotheses ................................................................................... 45

2.2.2.2.1. Tax considerations ................................................................................................... 46

2.2.2.2.2. Signalling hypothesis ............................................................................................... 47

2.2.2.2.3. Capital structure and corporate control .................................................................. 49

2.2.2.2.4. Behavioural finance ................................................................................................. 51

2.3. M&A performance ............................................................................................................ 55

2.3.1. Short-term performance ................................................................................................. 56

2.3.2. Long-term performance ................................................................................................. 60

2.3.2.1. Stock performance ...................................................................................................... 60

2.3.2.2. Operating performance ............................................................................................... 62

2.3.3. Factors influencing M&A performance ......................................................................... 65

2.3.3.1. The method of payment ............................................................................................... 65

2.3.3.2. Target status................................................................................................................ 68

2.3.3.3. Merger relatedness ..................................................................................................... 69

2.3.3.4. Merger attitudes .......................................................................................................... 70

2.3.3.5. Merger choice ............................................................................................................. 70

VI

2.3.3.6. Size and relative size ................................................................................................... 72

2.3.3.7. Market-to-book ratio ................................................................................................... 74

2.4. Prospect theory.................................................................................................................. 76

2.4.1. Diminishing sensitivity ................................................................................................... 77

2.4.2. Reference dependence .................................................................................................... 77

2.4.3. Loss-aversion tendency .................................................................................................. 78

2.4.4. Prospect theory and M&As ............................................................................................ 79

CHAPTER 3: REFERENCE POINT THEORY ON CROSS-BORDER AND DOMESTIC

M&A DEALS: U.K. EVIDENCE ........................................................................................... 81

ABSTRACT ............................................................................................................................. 82

3.1. INTRODUCTION ............................................................................................................ 83

3.2. LITERATURE REVIEW ................................................................................................. 90

3.2.1. Cross-border M&As in the U.K. .................................................................................... 90

3.2.2. The reference point effect on the U.K. ........................................................................... 93

3.2.3. The target 52-week high as a reference point price ...................................................... 96

3.2.4. M&A offer premiums ..................................................................................................... 98

3.2.5. Reference point price and the offer premium .............................................................. 103

3.3. HYPOTHESIS DEVELOPMENT.................................................................................. 105

3.4. DATA AND METHODOLOGY .................................................................................... 108

3.4.1. Data.............................................................................................................................. 108

3.4.2. Methodology ................................................................................................................ 114

3.4.2.1. Piecewise linear regression and local polynomial smooth procedure ..................... 114

3.4.2.2. Short-term event study .............................................................................................. 117

3.4.2.2.1. Market-adjusted model .......................................................................................... 117

3.4.2.2.2. Market Model ......................................................................................................... 119

3.4.2.3. Multivariate analysis ................................................................................................ 120

3.4.2.4. Two Stage Least Squares (2SLS) .............................................................................. 121

3.5. EMPIRICAL RESULTS ................................................................................................. 122

3.5.1. The reference point effect on the offer premium .......................................................... 122

3.5.2. The reference point effect on bidder announcement returns ....................................... 130

3.6. ROBUSTNESS CHECKS .............................................................................................. 132

3.6.1. The target 52-week high as a reference point .............................................................. 132

3.6.2. The reference point effect on the offer premium by market conditions ....................... 133

3.6.3. The reference point effect on the offer premium across different subsamples ............ 134

VII

3.6.4. The reference point effect on bidder announcement returns ....................................... 135

3.7. CONCLUSION ............................................................................................................... 136

CHAPTER 4: RELATIVE REFERENCE PRICES AND M&A MISVALUATIONS ........ 159

ABSTRACT ........................................................................................................................... 160

4.1. INTRODUCTION .......................................................................................................... 161

4.2. LITERATURE REVIEW ................................................................................................ 171

4.2.1. The misvaluation hypothesis ........................................................................................ 171

4.2.2. Reference point theory of M&As .................................................................................. 181

4.3. HYPOTHESIS DEVELOPMENT.................................................................................. 186

4.4. DATA AND METHODOLOGY ..................................................................................... 188

4.4.1. Data.............................................................................................................................. 188

4.4.2. Summary statistics ....................................................................................................... 192

4.4.3. Methodology ................................................................................................................ 193

4.4.3.1. Relative reference point (RRP) and offer premiums ................................................. 193

4.4.3.2. Logistic regressions .................................................................................................. 195

4.4.3.3. Classification of high-, neutral- and low-valuation markets .................................... 196

4.4.3.4. Short-term method .................................................................................................... 197

4.4.3.5. Long-term method ..................................................................................................... 197

4.4.3.5.1. Market-adjusted model .......................................................................................... 197

4.4.3.5.2. Size-adjusted model ............................................................................................... 198

4.4.3.5.3. Bootstrapped t-statistics......................................................................................... 199



4.5.1. The RRP effect on the probability of using stocks ........................................................ 201

4.5.2. The RRP effect on the offer premium ........................................................................... 203

4.5.3. Do bidders who focus on RRP overpay for the target? ............................................... 207

4.5.4. Do all stock-financed acquisitions driven by RRP protect the wealth of long-term

shareholders? ......................................................................................................................... 208


4.6.1. Endogeneity issues ....................................................................................................... 210

4.6.2. The effect of the RRP on the probability of using stocks.............................................. 211

4.6.3. The effect of the RRP on the offer premium ................................................................. 212

4.6.4. Do bidders who focus on RRP overpay for the target? Testing with the market-adjusted

model ...................................................................................................................................... 213

VIII

4.7. CONCLUSION ............................................................................................................... 213

CHAPTER 5: MARKET ANTICIPATION AND MERGER DECISIONS ......................... 236

ABSTRACT ........................................................................................................................... 237

5.1. INTRODUCTION .......................................................................................................... 238

5.2. LITERATURE REVIEW ............................................................................................... 245

5.2.1. Reference point theory and gambling .......................................................................... 245

5.2.2. Reference point theory and risk-taking ........................................................................ 248

5.2.3. Risk-taking and M&A outcomes .................................................................................. 253

5.3. HYPOTHESES DEVELOPMENT ................................................................................ 258

5.4. DATA AND METHODOLOGY .................................................................................... 260

5.4.1. Data.............................................................................................................................. 260

5.4.2. Methodology ................................................................................................................ 263

5.4.2.1. Target risks and the bidder 52-week high................................................................. 263

5.4.2.2. Short-term method .................................................................................................... 265

5.4.2.3. Long-term method ..................................................................................................... 265



5.5.1. Does market anticipation affect M&A decisions? ....................................................... 267

5.5.2. Do market-anticipated risks create value? .................................................................. 268

5.5.3. What is the source of bidder performance improvement? ........................................... 269

5.5.4. Do market-anticipated risks push the firm to work hard? ........................................... 270


5.6.1. Various reference point candidates based on firm level. ............................................. 271

5.6.2. Subsample tests ............................................................................................................ 273

5.6.3. Alternative proxies for target risks .............................................................................. 273

5.6.4. Risk-taking and M&A performance ............................................................................. 274

5.7. CONCLUSION ............................................................................................................... 274

CHAPTER 6: CONCLUSIONS ............................................................................................ 292

6.1. MAIN FINDINGS .......................................................................................................... 293

6.2. IMPLICATIONS, LIMITATIONS AND RECOMMENDATIONS FOR FUTURE

RESEARCH ........................................................................................................................... 298

REFERENCES ...................................................................................................................... 301

IX

List of Tables

Table 3.1: Summary statistics for M&A sample .................................................................... 138

Table 3.2: Distribution of bidder nations .............................................................................. 139

Table 3.3: Summary statistics for variables .......................................................................... 140

Table 3.4: Summary statistics for variables by cross-border and domestic subsamples ...... 141

Table 3.5: The piecewise linear regressions of offer premiums on the target 52-week high . 142

Table 3.6:The piecewise linear regression of offer premiums on the target 52-week high of

cross-border and domestic acquisitions into the United Kingdom ........................................ 143

Table 3.7: The OLS regression of bidder announcement returns on offer premiums ............ 144

Table 3.8: Robustness test: Controlling for other relevant price measures ........................... 145

Table 3.9: Robustness test: Controlling for target past returns ............................................. 146

Table 3.10: Robustness test: Reference point effect on offer premiums by market-wide

valuation ................................................................................................................................ 147

Table 3.11: Robustness test: U.K. and U.S. bidder subsamples............................................. 148

Table 3.12: Robustness test: Reference point effect on offer premiums by the method of

payment .................................................................................................................................. 149

Table 3.13: Robustness test: Reference point effect on offer premiums for diversified and

undiversified M&A deals ....................................................................................................... 150

Table 3.14: Robustness test: Reference point effect on offer premiums for tender offers and

mergers ................................................................................................................................... 151

Table 3.15: Robustness test: Market model for 5-day CARs ................................................. 152

Table 3.16: Robustness test: Market-adjusted model for 3-day CARs .................................. 153

Table 3.17: Robustness test: Market model for 3-day CARs ................................................. 154

Table 4.1: Summary statistics for M&A sample..................................................................... 215

Table 4.2: Summary statistics for variables ........................................................................... 216

Table 4.3: The effect of RRP on the probability of using stocks ............................................ 217

Table 4.4: The effect of RRP on the probability of using stocks under different market-wide

valuations ............................................................................................................................... 218

Table 4.5: The effect of RRP on offer premiums .................................................................... 220

Table 4.6: The effect of RRP on offer premiums over time .................................................... 221

Table 4.7: The effect of RRP on offer premiums by valuation periods .................................. 222

X

Table 4.8: Univariate analysis by different RRP groups ........................................................ 223

Table 4.9: OLS regressions of offer premiums on RRP: By RRP subsamples ....................... 224

Table 4.10: Do bidder focusing on RRP overpay for the target? .......................................... 225

Table 4.11: Do all stock-financed acquisitions driven by RRP protect the wealth of long-term

shareholders? ......................................................................................................................... 226

Table 4.12: Endogeneity issues .............................................................................................. 227

Table 4.13: Variables correlation matrix ............................................................................... 228

Table 4.14: Robustness check: RRP effect on the probability of using stocks ...................... 229

Table 4.15: Robustness test: The effect of RRP on the probability of using stocks: Between

payment methods .................................................................................................................... 230

Table 4.16: Robustness test: The effect of RRP on offer premiums: By payment methods ... 231

Table 4.17: Robustness test: The effect of RRP on offer premiums: By deal types ................ 232

Table 4.18: Robustness test: The effect of RRP on offer premiums: By deal choice ............. 233

Table 4.19: Robustness test: The effect of RRP on offer premiums: By deal completion ...... 234

Table 4.20: Robustness test: market-adjusted model ............................................................. 235

Table 5.1: Summary statistics for M&A sample..................................................................... 276

Table 5.2: Summary statistics for variables ........................................................................... 277

Table 5.3: Risky and less risky acquisition sample ................................................................ 279

Table 5.4: Reference point effect on M&A decisions ............................................................ 280

Table 5.5: OLS regressions of bidder CARs on market-anticipated risks ............................. 281

Table 5.6: Whether market-anticipated risks increase offer premiums? ............................... 282

Table 5.7: Do market-anticipated risks push the firm to work hard? .................................... 283

Table 5.8: Robustness tests: Various reference candidates in relation to the firm level ........ 284

Table 5.9: Robustness test: Do the market returns or peer pressures play a role in the firm’s

risk-taking? ............................................................................................................................ 286

Table 5.10: Robustness test: Subsample tests ........................................................................ 288

Table 5.11: Robustness test: OLS regressions of target idiosyncratic risk on the bidder

reference point ....................................................................................................................... 289

Table 5.12: Robustness test: OLS regressions of bidder CARs on market-anticipated risks . 290

Table 5.13: Robustness test: OLS regressions of BHARs on market-anticipated risks ......... 291

XI

List of Figures

Figure 3.1: Distribution of number of deals by bidders ........................................................ 155

Figure 3.2: Distribution of value of deals by bidder types .................................................... 156

Figure 3.3: Offer premiums relative to the target 52-week high ........................................... 157

Figure 3.4: Nonlinear relation between offer premiums and the target 52-week high .......... 158

XII

“We're blind to our blindness. We have very little idea of how little we know. We're not

designed to know how little we know.”

Daniel Kahneman

1

CHAPTER 1: INTRODUCTION

Chapter 1. Introduction

2

The central purpose of this thesis is to examine the reference point effect on the causes and

consequences of Mergers and Acquisitions (M&As). In Collins’ English Dictionary: the

reference point is ‘a fact forming the basis of evaluation or assessment’. The notion of the

reference point was first noted in the work of Tversky and Kahneman (1974), and developed

as an important implication of Kahneman and Tversky’s prospect theory (1979). Prospect

theory is a descriptive theory depicting how individuals make decisions under risks. A series

of implications were built in relation to the reference point, a specific point that gauges people’s

mental feelings about gains and losses. Prospect theory suggests that people have a strong loss-

aversion tendency relative to a reference point, suggesting that people are unwilling to suffer

losses and thus seeking many risks when experiencing loss, while, on the other hand, they avoid

taking too many risks when enjoying gain. Feelings about the loss are more intense than those

relating to gain, and feelings tend to be stronger nearer the reference point than further away

from it in both the loss and the gain domains.

M&As serve as an ideal testing ground for prospect theory for four proposed reasons. First,

M&As are major corporate investment activities associated with large risks. Prospect theory

addresses how managers make decisions in consideration of uncertainty, which is a suitable

theory to explain the M&A motive. M&As decisions are complex as they normally involve a

negotiation process that requires superior skills and considerable efforts by bidder managers

who are committed to completing the M&A deal to create the wealth for shareholders. It

remains largely unexplored how much a bidder manager should pay for the target. Valuing

targets is a very subjective task since it rests upon a large number of assumptions that make the

decision-making process hard to be observed by outsiders. In such a case, the reference point

price, a relevant piece of information, is likely to be employed to assess M&A motive.


3

Secondly, whilst existent behaviour finance theories have explained M&A motives on a basis

of deal synergies, the hubris management hypothesis (Roll, 1986; Hayward and Hambrick,

1997), managerial overconfidence (Malmendier and Tate 2005, 2008), self-attribution bias

(Doukas and Petmezas, 2007) and the misvaluation hypothesis (Shleifer and Vishny, 2003;

Dong et al., 2006), prospect theory explains the M&A motive from a perspective of managerial

decision-making process. This theory explains decision-making from a human’s nature

response, without making any assumptions regarding whether the market or the managers are

rational during the course of the transaction. Given that the outcome of major investment

decisions significantly affects the interests of stakeholders, the reference point price, as it is

easily obtained, is likely to be considered by the managers. The reference point theory of

M&As thus allows the researcher to observe how the reference-dependence bias affects

managerial major corporate investment decisions, such as initiating an M&A deal or valuing

the target.

Thirdly, whilst traditional theories explain the M&A motive from the perspective of the bidder,

reference point thinking could help the researcher to explain M&A motives from the

perspectives of the two firms involved in an M&A deal (i.e. the two firms involved). Because

the agreement of an M&A deal cannot be reached without the consent of the two firms involved,

bidders should consider the willingness for deal acceptance on the target side while making

M&A decisions. The reference point used in this thesis is an item of price information that

reflects the market’s perspective on a firm’s prospects. Managers who make major decisions

should also be concerned about the market’s reactions. Therefore, focusing on the reference

point effect enables a big picture on how investment strategy is formulated and how it is

structured from the perspectives of major participants in M&As. This thesis examines role of

the reference point effect played on various central aspects of M&As, including offer premiums,


4

the method of payment and managerial risk-taking attitude, which significantly affect M&A

outcomes.

Finally, M&As offer the researcher many considerations on how the reference point effect is

reinforced. One possible explanation for people looking at a salient piece of information is

because there exists considerable uncertainty about the information, making the true situation

less likely to be accurately observed. This raises the concern as to whether the reference point

effect is significant in the setting where the level of information asymmetry is high according

to the information asymmetry hypothesis. Addressing this concern, the reference point theory

of M&As was made use of in a sample consisting of both cross-border and domestic bidders’

acquisitions of a U.K. target. It was expected that the reference point effect would be stronger

for cross-border bidders than domestic bidders based on the rationale that domestic bidders,

with sophisticated networking, are more advantageous in gathering information than cross-

border bidders. A common intuition suggests that managers are skilful and have private

information about their firms, and have better understanding than the market in terms of the

deal. Thus, it is expected that managers (i.e. more experienced investors) should be less likely

to be influenced by a single number, such as a reference point price, instead, they tend to take

other relevant factors into consideration, whilst the market (i.e. less experienced investors)

tends to put greater emphasis on the reference point due to limited information and limited time

in which to process the deal. If this is the case, the reference point effect should have little

impact on managers’ decision-making process. However, it may not be the case, as managers

who work for the best interests of the shareholders should provide rationale to shareholders as

to why they initiate an M&A deal and how it is valued, implying that the reference point is

used as an important decision-making tool for shareholders to judge M&A performance. This


5

thesis integrating prospect theory into the M&A testing ground gives the researcher better

knowledge on how different M&A stakeholders understand M&A motives.

Neoclassical M&A theory suggests that managers are rational and thus add the value to the

firm through M&As (Jensen and Ruback, 1983). According to Jensen (1988), M&As are a

process of allocating and integrating resources of a firm so as to maximise the firm’s value.

His view is well supported by Berkovitch and Narayanan (1993) and Hodgkinson and

Partington (2008) who found synergies gains an essential motive for M&As, dominating over

other proposed M&A motives, such as hubris and agency problems. Despite this, it is a well-

documented fact that bidders do not gain in either the short or the long run, whilst targets gain

in M&As (Jensen and Ruback, 1983). The fundamental reason for bidder’s underperformance

is overpayment, which is the offer price paid to the target shareholders largely is greater than

the expected synergies gains following a merger. Bidder shareholders may find the firm has

difficulty generating synergies if high offer premiums are offered to the target shareholders,

since they are directly observed by the market, leading to negative market reactions once the

bid is announced.

One thread in relevant literature contributes the concept of overpayment to the agency conflicts

of interests between managers and shareholders, which arise when managers seize the

opportunity to acquire firms to pursue their own benefits at the expenses of their shareholders.

As a result, bidder managers tend to pay excessively for targets. Jensen (1986) studied the

performance of a bidder with large free cash flow and found that managers tend to engage in

ill-conceived investment projects. In a similar vein, Morck et al. (1990) find managers fail to

maximise the value of the firm, as they tend to engage in value-destroying deals for the

purposes of risk reduction and job security. Fu et al. (2013) found that severe agency problems


6

lead to the firm’s underperformance in the long term. All of these studies indicate that

managerial overpayment destroys the value of a firm.

Another thread in relevant literature concerns overpayment to behavioural finance explanations.

Roll (1986) was the first to develop the hubris management hypothesis that managers tend to

overestimate the synergies gains to be generated by M&As and accordingly overpay for the

target. In this line of thinking, Hayward and Hambrick (1997) found the self-importance of top

managers leads to overpayment. Malmendier and Tate (2008) found overconfidence in

managers having a strong preference for their firm’s stocks tends to lead to the completion of

more acquisitions associated with negative market reactions. Rau and Vermaelen (1998)

contribute such managerial overconfidence to market-to-book value (MTBV), showing that

high MTBV firms tend to be associated with strong stock price momentum due to better

historical performance, thus being overvalued. The positive market reaction therefore

reinforces the managerial illusion of control1, leading to excessive unnecessary acquisitions.

Moeller et al. (2004) ascribe managerial confidence to firm size, suggesting that larger firms

are associated with worse performance than that of smaller firms, reasoning that managers of

larger firms tend to have fewer restrictions on a firm’s resources and such power of autonomy

arguments managerial overconfidence.

Prospect theory offers many additional considerations for overpayment distinct from those in

behaviour finance theories. Given limited time and limited information in which to process the

deal, bidders and targets tend to employ a relevant piece of price information for decision-

making. Following studies that link corporate investment decisions and the firm’s reference

point price (Heath et al., 1999, George and Hwang, 2004, Huddart et al., 2009, Burghof and

1 The tendency that people overestimate their ability to control events.


7

Prothmann, 2011, Baker et al., 2012), the firm’s 52-week highs were employed as reference

points in this thesis, as they are frequently reported by financial media and more likely to create

a deep impression in the investors’ minds. Targets may presumably believe that a price relative

to their 52-week high is a fair price as it reflects their best performance. According to prospect

theory, people have a strong loss-aversion tendency. If losing the control of the firm is

perceived as a mental loss for the target shareholders, an offer price that is closer to the

reference point is more likely to compensate the feelings of such a loss, making the deal more

acceptable. On the other hand, targets encouraged to use its 52-week high to negotiate a decent

offer price with the bidder, as target shareholders believe the firm may well create value without

it being acquired. Edward and Walkling (1985) regard offer premiums as the outcome of the

bargaining process between the bidder and target managements, suggesting that bidders in a

relatively weak position tend to pay high offer premiums, leading lower abnormal returns,

while targets in a relative strong position can demand high offer premiums being translated

into larger positive abnormal returns. Therefore, the use of the reference point reinforces the

target’s bargaining power in exchange for high M&A offer premiums.

Chapter 3 of this thesis examines the role of reference point effect in the U.K. market. The idea

of Baker et al. (2012), who examined the role of the target 52-week high effect on the U.S.

market, was put forward to the U.K. market. Unlike Baker et al. (2012), a sample containing

both the domestic and cross-border bidders were taken into consideration in the United

Kingdom. The reference point effect was expected to have a substantial impact on cross-border

M&As into the U.K. for three proposed reasons. First, U.K. targets may find it easier to use

their 52-week high to negotiate with the bidder given that targets are in the most competitive

market where their bargaining power is more easily to be reinforced and thus being translated

into higher offer premiums (Moeller and Schlingemann, 2005; Alexandridis et al., 2010).


8

Second, U.K. firms have a relatively strong corporate governance system where target

managers are expected to be more likely to serve the best interests of their shareholders. In that

case, target managers tend to use their 52-week high to seek higher offer premiums, which

reconcile the view of their shareholders. Finally, it is expected a stronger tendency for cross-

border bidders than domestic bidders in terms of looking at the target 52-week high as a

reference point given that there is greater information uncertainty regarding cross-border

bidders over domestic bidders.

Using a sample of 451 domestic acquisitions and 155 cross-border acquisitions into the U.K.

between 1985 and 2014, a positive relationship between the target 52-week high and the offer

premium was emerged. The reference point distinguishes the takeover motive of domestic

bidders from that of cross-border bidders. It became evident that offer premiums driven by the

target 52-week high lead to a negative market reaction for domestic bidders, suggesting

evidence of overpayment. However, there is little evidence of overpayment for cross-border

bidders paying according to the target 52-week high. The findings in this research are consistent

with prospect theory.

Chapter 3 makes two distinct contributions to the M&A literature. First, this is the first

examination of the role of the reference point in the context of cross-border M&As. This allows

the researcher to investigate how bidders value targets under information uncertainty. Further,

the reference point theory of M&As distinguishes the motive of domestic bidders from cross-

border bidders acquiring U.K. public targets. Offer premiums paid based on the target reference

point price have a negative impact on bidder announcement returns for the domestic M&A

subsample while there is an insignificant impact on bidder announcement returns for the cross-

border M&A subsample, indicating that shareholders tend to believe that managers of domestic


9

bidders pay too much for the target based on the target 52-week high, while take those of the

cross-border who pay according to the target 52-week high for granted.

Chapter 3 applies the idea of the reference point effect to the U.K. where targets have a strong

bargaining position with an expectation that they are likely to employ the 52-week high as the

reference point in the negotiation table. This chapter finds that the reference point reinforces

the target’s bargaining power in M&As, and important factors that enhance the reference point

effect on M&As. However, it leaves a great deal to think about the role of the bidder reference

point in the context of M&As in consideration of the bidder 52-week high is also an available

item of information for the market to assess the value of the firm around the takeover

announcement date. Whilst Chapter 3 finds that the target 52-week high is a reference point for

how much the bidder should pay for the target, Chapter 4 studies the bidder 52-week high effect,

which gives market information on how much the bidder could pay for the target. Bidder with

a nearness 52-week high tend to create an image of strong profit-generating ability and richness

in financial resources, whilst those whose current price deviates considerably from the 52-week

high tend to experience many difficulties to finance a deal.

In addition, Baker et al. (2012) reason that bidders who pay according to the target 52-week

high are confident in realising synergies above the same level of the target. However, these

authors found that those bidders do not generate synergies. If they can time the market by

selecting quality deals for the purpose of synergies integration, this leaves the researcher of this

thesis to believe that focusing on the target 52-week high alone may not fully account for the

reference point effect on M&As. Therefore, both the target and the bidder 52-week highs were

taken into considerations and examined in Chapter 4, where a new proxy was developed by

including both the target and the bidder reference points. If the market price reflects the view


10

of the market investors to the firm’s valuation, it should be suggested that a firm with the

nearness 52-week high is likely to be overvalued based on the rationale that investors should

be reluctant to bid up the price, fearing that the firm is already overvalued. In contrast, a firm

whose current price deviates a long way from its 52-week high leads the market to believe that

the firm is experiencing bad news and is thus undervalued. Therefore, prospect theory can

accommodate many important implications of the misvaluation hypothesis.

Whilst behaviour finance theories explain value-destroying deals as bidders have a better than

average illusion and overpay the target shareholders accordingly, Shleifer and Vishny (2003)

proposed the misvaluation hypothesis that bidders overpay for the targets with the purpose of

diluting overvalued stocks. This hypothesis rests upon the assumption that the market is

inefficient whereas managers are rational, thus managers are able to exploit mispricing of the

market. However, this motive is not observed by investors with short horizons, resulting in

negative market reactions to the bid announcement. The theoretical model of the misvaluation

hypothesis yields three important implications: first, overvalued firms tend to pay high offer

premiums to the shareholders of a less overvalued firm. Second, overvalued firms prefer stocks

as a means of finance for acquisitions as it accelerates the process of overvaluation dilution.

Finally, offer premiums hedge against the risks of holding stocks in an overvalued market.

Studies reveal that the rationale behind overpayment is that bidders are threatened by the

danger of overvaluation, as addressed in the Jensen’s work (2005).

Shleifer and Vishny’s study (2003) fail to provide any rationale as to why targets would accept

overvalued stocks given that they are as rational as bidders and their firm is not

underperforming prior to an M&A deal. Rhodes-Kropf and Viswanathan (2004) complemented

their view by arguing that both bidders and targets are rational and the reason for making


11

mistakes by accepting too many overvalued stocks is that targets do not have access to private

information about the bidder, resulting in large valuation errors due to an overestimation of

synergies. Rhodes-Kropf et al. (2005), further relaxing the irrational target assumption,

suggesting that targets will also make mistakes in valuing the deal when market-wide valuation

is high. Dong et al. (2006), who employed the P/B and the P/V of both the bidder and target to

examine what drives M&As, also provided supportive evidence that estimation biases of

managers to the firm’s valuation leads to M&As. They note that it is not the true market

misvaluation of the firm but the view of managerial misvaluation drives M&As.

However, Di Guili (2013) challenges this view by proposing the measures for misvaluation.

He found that a high MTBV could lead to predictions that the firm has better investment

opportunity or the firm is significantly overvalued, leading the researcher to believe previous

studies borrowed such valuation measure cannot fully account for the misvaluation hypothesis.

So as to overcome the bias caused by the firm’s fundamental characteristics, Chapter 4 of this

thesis measured the misvaluation of the firms involved by constructing a new proxy, the

relative reference point (RRP), which is the difference between the target and the bidder’s

reference points. The measure makes no assumption as to the true valuation of the firms but

aims to explain the valuation errors driven by the market affecting M&As. It could be advisable

that market should hold a similar view of the valuation of the two firms involved prior to the

announcement date. The measure covers considerations of both the target and bidder sides.

Since the market price is a reflection of the investors’ view of the firm performance, the RRP

thus directly reflects the market’s perception of a firm’s misvaluation. The measure built upon

the reference point captures market reactions and eliminates estimation biases caused by a

firm’s fundamental characteristics.


12

Using a sample of 1,878 U.S. domestic public M&As between 1985 and 2014, it became

apparent that the propensity for using stocks increases when the RRP increases, suggesting that

the relatively more overvalued bidders tend to use stocks as a means of finance for acquisitions.

In addition, the offer premium increases with the RRP, indicating relatively more overvalued

bidders tend to pay higher offer premiums to the less overvalued targets. The RRP takes on a

role of overpayment in the short term, reflected in negative market reactions. However, the

long-term analysis indicates that offer premiums that are associated with acquisitions of higher

RRP tend to generate less negative long-term returns than acquisitions of a lower RRP,

suggesting that managers of more overvalued firms are able to time the market by looking at

the RRP.

Chapter 4 provides three distinct contributions to the M&A literature. First, a new measure for

misvaluation is proposed, based on market perception. The measure establishes a bilateral

valuation framework where both the bidder and the target’s misvaluation are considered. The

measure also eliminates the biases caused by a firm’s fundamental characteristics or any

retrospective and forward-looking information since it reflects the latest market reaction.

Secondly, we provide evidence that both experienced and less experienced investors tend to be

affected by the reference point effect. However, they may interpret the reference point in

different ways. An investor with a short-term horizon tends to contribute the offer premium

paid to the target shareholders with the result of overpayment, whilst a bidder who pay offer

premiums according to the RRP times the market, reflected in the less negative market. Finally,

it is suggested that the RRP is an important indicator of the method of payment. Overvalued

bidders who use stocks rather than cash as a means of finance for acquisitions tend to pay lower

offer premiums. It was established that lower offer premiums are associated with cash instead


13

of stock payments for relatively more undervalued targets. These results offer a new

explanation for the method of payment hypothesis.

One important implication of the reference point is that people tend to seek risks when they

experience loss relative to the reference point and adopt a risk aversion when they enjoy gains

relative to this specific point. Chapter 5 of this thesis directly assessed this implication by using

the bidder 52-week high as the reference point. Since the firm’s 52-week high reflects the

market reaction to the bidder’s performance, the loss relative to a reference point price serves

as a relevant piece of information that signals to the market that risky investment opportunities

should be available. Managers who fail to utilise this will be blamed for not meeting the

commitment of value-maximising the firm. Therefore, a decline in performance prompts

managers to undertake risky M&As. Managers are also exposed to larger risks when they

engage in a risky M&A deal, motivating them to rationalise their motive by taking market-

anticipated risks and working hard to convince the market the performance will be boosted

through their superior management skills in risky projects. It is expected a firm’s current price

that deviates greatly from its 52-week highest stock price leads managers to undertake M&As

according to market anticipation. As a result, risky M&As would bring forth positive abnormal

returns for the firm. Managers engaging in risky acquisitions are also exposed in stricter

shareholder monitoring, requiring them to make greater efforts on deal negotiation, reflected

in the lower offer premiums they pay for the target. In a further analysis, the quality of the

decision that is based on market anticipation is assessed. Since the integration process of M&As

requires time, it is suggested that assessing the short-term performance may not fully reflect

the quality of a risky M&A decision. Therefore, the long-term performance of the deal is

studied to observe whether bidder managers make great efforts to improve the deal quality.


14

M&As form is a good platform to test managerial risk-seeking behaviour. First, they are a

major investment activity that could fundamentally alter the wealth status of the shareholders,

since small magnitude of change in wealth status does not touch investors’ risk appetizer

(Brenner, 1983). Further, it is common for shareholders to compensate managers engaging in

risky investment projects (Grinstein and Hribar, 2004; Graham et al. 2013). The feeling of loss,

part of human nature, tends to motivate the manager to tolerate more risks and accordingly

undertake risky M&As. Finally, M&As allow the researcher of this thesis to further explore

whether managers can control risks with their skills and efforts. It was also found that risk-

taking is distinguished from gambling insofar as it depends purely on probability.

Chapter 5 makes two main findings by studying a sample of 2,018 U.S. public acquisitions

announced between 1985 and 2014. First, the bidder reference point and target risks are

positively related, implying that when bidders perceive losses relating to the reference point

they are likely to undertake risky investment projects. Second, bidders engaging in risky

projects anticipated by the market tend to perform better, suggesting that managers who follow

market-anticipated risks will be rewarded. Third, it was also established that risky M&As

expose managers to strict shareholder monitoring, which motivates them to work hard.

Chapter 5 makes three distinct contributions to M&A literature. First, this is the first paper to

test the reference point effect on managerial risk-taking behaviour in the M&A context. It is

also suggested that human-related biases have a direct impact on investment behaviour. Second,

the researcher contributes to behaviour finance literature by documenting the evidence of the

reference point effect on both individual and institutional investors. Third, a contribution is

made to M&A literature, affirming that taking risks does not necessarily lead to negative


15

market reactions and taking risks according to market’s anticipation is an important source of

positive market reactions.

The remainder of this thesis is organised as follows: Chapter 2 provides a summary of relevant

literature on M&A motives, M&A processes, M&A performance and prospect theory. Chapter

3 studies the reference point theory in domestic and cross-border M&As in the U.K. market.

Chapter 4 examines the reference point effect on M&A misvaluations, while Chapter 5 assesses

the reference point effect on managerial risk-taking behaviour. Chapter 6 concludes this thesis.

16

CHAPTER 2: LITERATURE REVIEW

Chapter 2. Literature Review: M&A Motives

17

2.1. M&A motives

Researchers on mergers and acquisitions (M&As)2 focus on the causes and consequences.

M&As are one of the major investment activities undertaken by firms to create wealth for

shareholders, expand the size, or integrate resources for example. Prior studies have offered

various explanations for M&A motives. So as to prescribe corresponding strategies for

management following M&As, Trautwein (1990) recognised seven M&A motives, namely

efficiency theory, monopoly theory, valuation theory, empire-building theory, raider theory,

process theory and disturbance theory. Of these theories, he finds that the valuation theory,

empire-building theory and process theory play a more active role in M&A motives than the

efficiency theory and monopoly theory, whereas raider theory and disturbance theory have the

least explanatory power. Despite this, his study implies that no single theory can fully account

for an M&A motive.

Studies have also provided empirical evidence in explaining the M&A motives. Berkovitch

and Narayanan (1993) were the first to test the fitness of the three major M&A motives: synergy,

agency and hubris. The authors studied a sample of tender offers in the U.S. market between

1963 and 1988 and used the correlation between the gains of the target and the combination to

examine M&A motives. They assumed that the positive correlation between gains of the target

and the combination represented synergy, the negative correlation represented agency whilst

the zero correlation represented hubris. Their study indicates that about 75 percent of M&As

are motivated by synergy and the negative correlations are driven by agency, while hubris co-

exists with synergy. Hodgkinson and Partington (2008) put forward their idea in an alternative

market. They examined a sample of 529 successful M&A deals between 1984 and 1998 in the

2 In this thesis, the terms ‘takeovers’ and ‘mergers and acquisitions’ were used interchangeably, ‘mergers’ and

‘acquisitions’ also refer to M&As in some circumstances unless they are specified in the correspondent sessions.


18

U.K. market and found that synergy was the most plausible M&A motive and there is also

evidence for both agency and hubris. Raj and Forsyth (2003), who focused on the performance

of hubris-infected bidder managers, noted significant loss on the bidder announcement returns,

suggesting that one of the negative sources of bidder performance is managers’ overestimation

of M&A synergies.

Seth et al. (2000), who studied the bidders’ motives in the context of cross-border M&As,

affirm that synergy remains a primary motive for cross-border M&As. Studying 100

acquisitions of U.S. targets between 1981 and 1990, the authors found synergy is the

dominating M&A motive for domestic M&As consistent with the findings of Berkovitch and

Narayanan (1993). In addition, the total gains in cross-border M&As are 7.6 percent, which is

similar to the figure achieved in the study conducted by Bradly et al. (1988)3. Seth et al. (2002)

identified multiple sources of value creation and value destruction for cross-border M&As.

Based on their sample, asset sharing, reverse internationalisation of valuable intangible assets

and financial diversification are the main sources of value creation while risk reduction is the

only source for value destruction. Their paper suggests that synergies play an essential role in

cross-border M&As.

2.1.1. Synergies

The synergy hypothesis of M&As asserts that the value of combined firms after a merger is

larger than the sum of the returns generated by the two separated firms. Bradley et al. (1983)

found some support for this hypothesis, suggesting that M&As create synergies for the

combined entity by reallocating the resources of the firms. Unlike the information hypothesis,

which posits that managers have private information about the true value of the firm, authors

3 Bradly et al. (1988) found a 7.4 percent increase of the combined firm in domestic M&A sample.


19

suggest that managers generate gains by exploring potential synergies via M&As, which is

consistent with the synergy hypothesis. Bradly et al. (1988) investigated a sample of 236

successful tender offers in the period 1963 to 1984 and recorded an overall increase of 7.4

percent in the value of combined firms, providing additional support to the synergy hypothesis.

Jensen and Ruback (1983), by reviewing prior studies, found that M&As on average generate

gains, and targets benefit, whereas bidder firms at least do not lose.

A large body of M&A studies have summarised three main sources of synergies: operations,

finances and management improvement. Devos et al. (2009) attribute synergies to an

improvement of resources allocation. Using the value line forecasts to estimate synergy4, the

authors found that the average synergy for a sample of 246 large mergers increased the value

of the combined entity to 10.03 percent. Of particular note, operating synergies take up 8.38

percent and financial synergies account for the remaining part. Hoberg and Phillips (2010)

suggest that large synergies are obtained through acquisitions of targets whose products are

different from bidder firms’ rivals, suggesting that synergies are created by a product

differentiation strategy that intensifies bidders’ competitiveness and sets obstacles for their

rivals, making them hard to imitate. Using a large unique patent-merger dataset between 1984

and 2006, Bena and Li (2014) found synergies to be created through innovation activities

following M&As. Houston et al. (2001), who investigated a sample of the largest bank mergers

between 1985 and 1996, noted that bank mergers create value for bidders. In particular, the

authors identified the source of synergy in bank M&As as being mainly created from cost

savings rather than revenue enhancement.

4 In Devos et al. (2009), merger synergy was calculated as the forecasted incremental cash flows of the combined

firms relative to the sum of the premerger forecasted cash flows of bidding and target firms.


20

Lang et al. (1989) explained synergy gains with the Q theory. Tobin’s Q is defined as the ratio

of a firm’s market value to the replacement costs of its assets. Q ratio measures the firm’s

performance, and the well-performing firms are those with a high Q ratio whereas the poor-

performing firms are those with a low Q ratio. Acquisitions involve a high Q bidder and a low

Q target generating large total gains for the bidder, the target and the combined firm. The results

imply that managers of a well-performing firm have abundant resources to improve the firm’s

performance whereas a poor-performing firm lacking investment opportunities is an ideal

target for synergy exploitation. Servaes (1991) reported supportive evidence by applying Lang

et al.’s idea (1989) to a larger sample with additional controls.

M&As play a disciplinary role with regard to managers who perform poorly. Jensen and

Ruback (1983) indicate that the market for corporate control creates value for the firm.

Managers compete for the rights to manage corporate resources, hence, poor management is

removed while good management serves the best interests of the shareholders via M&As.

Jensen and Ruback’s view (1983) suggests that M&As enhance the firm’s efficiency by

removing an inefficiency management team. Palepu (1986) studied the inefficient management

effect on the likelihood of the firm to be acquired, and measured an inefficient management

team by means of average excess returns of the firm. His results show a negative relationship

between firms that are likely to be the M&A target and inefficient management, which is

significant at a 5% level. The conclusion of this inefficiency management hypothesis is that

managers who fail to ensure the commitment of value-maximisation for the firm are likely to

be replaced. Accordingly, Mitchell and Lehn (1990) found that firms whose managers perform

poorly following a merger is likely to be acquired by another firm. Martin and McConnell

(1991) affirmed high turnover rate for target managers following M&A completion and the

performance of these firms is well below that of the industry average prior to the M&A. In a


21

similar vein, Lehn and Zhao (2006), by studying 714 acquisitions during 1990 to 1998, found

that 47% of top managers of bidder firms are dismissed within 5 years (including 16% due to

takeovers). Therefore, takeover threats for managers tend to create synergy for the firm.

However, Agrawal and Jaffe (2003), who included a sample of over 2,000 M&As from 1926

to 1996, reported little evidence that inefficient firms will necessarily be acquired. Measuring

the operating performance and stock market returns for the target firms, the authors revealed

that takeover targets are not all those underperforming firms prior to the M&A announcement.

In addition, targets do not underperform in the decade prior to M&A announcements regardless

of whether the performance was measured with the firm’s operating performance or stock

market performance. Results are robust when the authors take account of the size of the firm,

industry and past performance of the targets for the targets’ operating returns, and the size,

book-to-market value and past returns for the stock market performance. Results continue to

hold when employing an alternative benchmark model to calculate targets’ stock returns. The

authors proposed two main reasons why their findings fail to support Palepu’s inefficient

management hypothesis (1986). First, the primary M&A motive is not to remove the inefficient

management team in the sample, as there is only a small proportion of targets perform

inefficiently prior to M&A announcement date. Second, their sample did not include

disciplinary or attempted M&As that support the inefficient management hypothesis.

Jensen (1988) suggests that the market for corporate control brings forth economic benefits for

shareholders, society and the corporate form of organisation. Moreover, takeovers bring about

major changes fitting for the future development of firms. Such fundamental changes including

new recruitment and resources, offer new top-managers opportunities, forsaking the old

strategies and facilities that are no longer beneficial to the firm. Further, the market for


22

corporate control effectively reallocates resources. Jensen (1988) indicates that major

restructuring activities create economic efficiency due to the influences of political and

economic conditions in the 1980s when many U.S. firms experienced revenue slowdown and

used M&As to cope with market changes. The author holds that the pressures of managers’

turnover and the political activity weaken M&A efficiency.

2.1.2. Agency problems

The firm is a separation of ownership and control, in which shareholders are the owners of the

firm (i.e. the principal) and managers are the agent hired by shareholders to manage the firm

(i.e. the agency). According to Jensen and Meckling (1976), this principal-agency relationship

is a financing contract between shareholders and managers, in which managers are committed

to maximising the wealth of shareholders. However, such a contract is violated when managers

pursue their own interests at the expenses of their shareholders (i.e. agency problems).

M&A studies have found evidence of agency problems in firms with sizable free cash flows.

Jensen (1986) suggests in the free cash flow hypothesis that managers of cash-rich firms tend

to distribute the excess of cash in unnecessary investment projects rather than paying out to

their shareholders. Harford (1999) provided empirical evidence in support of Jensen’s view.

Studying a large sample of U.S. M&As from 1950 to 1994, Harford found that managers of

cash-rich firms engaging in M&As due to low managerial ownership is where agency problems

are likely to occur. Cash-rich bidders receive negative market reactions and experience declines

in operating performance following M&As.

Contrary to the free cash flow hypothesis, Gregory (2005), focusing on the long-term

performance of U.K. market, noted that firms with abundant cash flows free of use outperform


23

those possessing low free cash flows. The author argued that firms with low free cash flows

are more likely to be financially constrained in the long term, and have difficulty in boosting

its performance following a merger. Managers who hold large free cash flows and whose firms

have a low Q ratio are likely to create value for the firm, in that the initial undervaluation tends

to be corrected by the market in the long term. Likewise, Lin and Lin (2014), using a sample

of Australia takeovers and measuring free cash flows with two proxies: excess cash holding

and excess accounting cash flow5, found that managers of cash-rich firms do not undertake

value-destroying M&As.

Morck et al. (1990) found “bad managers”6 who raise severe agency conflicts in M&As. Using

a sample of 326 U.S. M&As from 1975 to 1987, the authors found evidence of agency problems

through three types of value-destroying deals. First, agency problems are raised when managers

act for the purposes of job security and risk reduction of their own stock holdings. As such,

managers tend to undertake diversifying M&As associated with negative market reactions due

to unfamiliarity with an unrelated industry. Further, it is also suggested that bidders are likely

to overpay for high-growth targets to increase their personal benefits. Finally, firms performing

poorly in the past are likely to initiate bad deals. Harford et al. (2012) found entrenched

managers were less likely to pay for acquisitions with all-equity offers, avoiding creating larger

block-holders that diminish their power in the firm. Their M&A selection avoids acquiring

private targets that could create value for the firm7, and more importantly, they overpay for

targets whose synergy-generating ability is low.

5 Excess accounting cash flow was defined as the ratio of earnings after interest paid, tax paid and dividend paid

before depreciation to total assets. 6 “Bad managers” here refers to those who do not serve the best interests of their shareholders. 7 In Fuller et al. (2002), bidders’ acquisitions of private firms tend to generate positive announcement returns

while acquisitions of public firms generate negative announcement returns. Specifically, it is found that there is

negative 1 percent announcement returns when targets are public firms and positive 2.1 percent when targets are

private firms. In addition, Fuller et al. (2002) showed that offer premiums to private targets are lower than public

targets, which is also consistent with the finding in a study of French M&As over the period of 1966 and 1982 by

Eckbo and Longohr (1989).


24

Studies have shown one of the main sources for agency problems is the compensation

rewarded to top managers engaging in M&As. Studying a large sample of U.S. M&As, Harford

and Li (2007) found it was a great incentive for a bidder CEO to undertake an acquisition as

CEOs’ pay increases substantially after an acquisition regardless of the wealth of shareholders.

This is also supported by Guest (2009) who studied a comprehensive U.K. sample between

1984 and 2001. He established that CEOs’ salaries increase within a year after an acquisition

even though takeovers destroy the firm’s value.

Grinstein and Hribar (2004) revealed that it is managerial power that influences compensations

but not the deal performance, offering little evidence that better M&A performance increases

pay to the managers. Their study showed that 39% of companies reward their CEOs involved

in a completed deal due to the efforts the CEOs made, where efforts of those in the important

positions are easily to be seen. In contrast, Datta et al. (2001) and Falato (2008) found that

CEOs’ compensation is positively related to stock performance, indicating that compensation

boosts the firm’s performance, implying that there is low probability that compensation causes

agency problems.

Studies have also documented that not only bidder managers but also target managers could

cause agency problems, reflected in that target managers exchange private benefits such as

compensation or positions in the bidder firms following a merger for lower M&A offer

premiums. Analysing 311 large U.S. firms between 1995 and 1997, Hartzell et al. (2004) found

that target managers are likely to persuade shareholders to give up their control of a firm when

they are compensated with financial rewards or attractive positions in the bidder firms.


25

Likewise, Wulf (2004) in a sample of “mergers of equals”8 in the U.S. market between 1991

and 1999 found that target managers exchange shared governance in the bidder firms for lower

M&A offer premiums, which destroys the wealth of target shareholders, leading to 5.6 percent

lower than average abnormal returns. However, Bargeron et al. (2010) argued that target

managers’ retention is not as the result of lower M&A premiums. Their findings contradict

those of Hartzell et al. (2004) and Wulf (2004), indicating that bidders retain targets’ managers

who have skills and experience in the new combined firm to increase its value.

2.1.3. Behavioural finance theories

Behavioural finance has received a great deal of attention in M&As over the last three decades.

There are two threads in the literature that link behavioural finance with M&As, according to

a survey conducted by Baker and Wurgler (2012). The first thread of literature assumes that

the market is efficient while the managers are less than fully rational, in that managers cannot

benefit shareholders via M&As because there are no mispricing opportunities waiting to be

explored., and the hubris management hypothesis and the managerial overconfidence

hypothesis are the two of the most important hypotheses built upon this assumption. The second

thread of literature assumes that the market is inefficient while the managers are fully rational,

in that managers can explore market mispricing, and the over-reaction (or under-reaction)

hypothesis and the market-timing hypothesis are the two of the most important hypotheses built

upon this assumption.

8 According to Wulf (2004), “mergers of equals” was defined as that the bidder firm and the target firm is similar

in size.


26

2.1.3.1. Managerial overconfidence

Roll’s study (1986) was the first on managerial hubris in the context of M&As. The author

proposed his hubris management hypothesis as managers overpaying for targets due to an

overestimation to their own profit-generating ability. By reviewing U.S. and U.K. studies, Roll

(1986) identified overconfident managers and discussed them from three perspectives. First,

bidder managers are subject to hubris bias, in that the firm’s price falls after that a bid is

announced or fails at a later stage. Second, hubris infects target managers, as a target firm’s

price increases by 7 percent on average when a bid is announced and falls significantly when

the bid is withdrawn. Third, the performance of the combination entity is also affected by hubris

as it earns significantly negative abnormal returns following a merger. The hubris management

hypothesis provides evidence that an individual decision maker may not be rational, which

answers the question of why managers tend to pay for a takeover target with a price that is

higher than its market price.

Following hubris management hypothesis of Roll (1986), Hayward and Hambrick (1997), who

studied a sample of 106 large U.S. acquisitions, found that bidder CEOs paying too much for

large acquisitions is due to hubris bias. The authors identified three important sources of

managerial hubris: the firm’s recent performance, the recent media praise for CEOs’

performance and the CEOs’ self-importance. The authors suggest that any recent success of

the firm would enhance the confidence of the bidder managers. As a result, they overpay for

the targets. The media praise for CEOs also fosters CEOs’ overconfidence in investment

activities. Moreover, managers receiving with higher-than-average compensation and salary

tend to feel important in major investment decisions, which will make them believe their

abilities could improve the firm, thus overpaying for the targets. Similar to those of Roll (1986),

Hayward and Hambrick found hubris results in losses on bidder abnormal returns.


27

Malmendier and Tate (2005) revealed CEO overconfidence in a sample of 477 publicly traded

U.S. firms from 1980 to 1994. Measuring managerial overconfidence in the exercise of

managerial options and CEO stock holdings, the authors found that overconfident CEOs tend

to exercise their stock options at a later period and hold losers for a longer period. It is because

that they are over-optimistic about their firm’s prospects or have strong affiliation with their

own stocks, which miss the correct market timing for investments. Besides, their study also

revealed that CEOs whose firms have a higher investment to free cash flow ratio tend to

undertake more investments, suggesting that firms with overconfident managers are likely to

have sufficient internal funds. Using the same dataset as in the Malmendier and Tate’s work

(2005) and adopting some press-based proxies for managerial overconfidence, Malmendier and

Tate (2008) found that the market reacts negatively to a bid announced by overconfident CEOs

and affirmed that managerial overconfidence encourages more acquisitions since

overconfident CEOs are more aggressive than non-overconfident CEOs. These two related

studies lead the researcher to believe that small firms are less vulnerable to managerial

overconfidence while large firms tend to make more value-destroying deals due to

overconfidence.

Many studies have addressed the fundamental reason for managerial overconfidence. One

thread in M&A studies attributes managerial overconfidence to the market’s tendency to

explore the past best performance of big firms. Rau and Vermaelen (1998) put forward the

over-extrapolation hypothesis that the financial market and the managers tend to over-

extrapolate the past performance of the bidder and react more positively to those with low

book-to-market ratio (i.e. glamour firms) than firms with high book-to-market ratio (i.e. value

firms). The market’s reactions reinforce the confidence of the managers of glamour firms when

undertaking M&As. In addition, the authors found glamour firms perform better than value


28

firms in the short term as the market tends to reward firms with better past performance.

Making adjustments for the size of the firm and book-to-market value to estimate the long-term

performance of the bidder, the authors found that glamour bidders underperform value bidders.

They interpreted this phenomenon as that short-term momentum created by the market’s over-

extrapolation leads to correspondent long-term reversals when the market can correct its initial

reactions, and such a short-term momentum is stronger in glamour bidders than value bidders.

Sudarsanam and Mahate (2003) found consistent results when putting forward Rau and

Vermaelen’s idea (1998) for the U.K. market. They employed the price-to-earnings ratio (P/E)

and market-to-book value (MTBV) for glamour and value bidders and calculated shareholder

returns with a range of benchmark models. Their findings suggest that a lower P/E bidder (a

value bidder) outperforms a higher P/E bidder (a glamour bidder) and a lower MTBV firm

outperforms a higher MTBV firm in the long term.

Another thread in relevant literature has largely attributed managerial overconfidence to

managerial self-attribution bias. Attribution bias was first discussed in Heider’s attribution

theory (1958) in which suggests that people tend to make systematic errors and attempt to

attribute those to external factors. Doukas and Petmezas (2007) tested this theory on an M&A

testing ground. Studying a sample of 5,334 successful U.K. private acquisitions between 1980

and 2004, the authors recorded that bidders in the higher order acquisition9 underperform those

in the lower order acquisition. They describe this phenomenon as the self-attribution bias of

managers who believe initial success is due to their own superior ability and so undertake more

M&A deals, whilst managers attribute negative market reactions of the subsequent deals as bad

luck. Billet and Qian (2008) provided some similar findings by focusing on the U.S. market.

9 According to Doukas and Petmezas (2007), higher order acquisitions refer to bidders who undertake five or more

deals within a three-year period.


29

After controlling for the order of the deal, their findings showed that the first bids have higher

abnormal returns whilst the higher order bids generate negative abnormal returns. In addition,

bidders who have experience of undertaking successful acquisitions before are likely to engage

in more value-destroying M&A deals. It is suggested that bidder managers are cautious in their

first bids but are affected by self-attribution bias when undertaking subsequent bids.

However, Aktas et al. (2009) explain the findings of Doukas and Petmezas (2007) and Billet

and Qian (2008) as the learning hypothesis. They suggest that the negative returns to the bidder

firm during the time of a bid announcement are a result of the learning behaviour of bidder

managers. The authors reason that bidder managers who undertake multiple bids tend to learn

lessons from the previous deal and become more risk-averse for subsequent deals, which leads

to negative abnormal returns. This study links a positive relationship between market reactions

and managerial risk-taking. It is worth noting that the learning hypothesis weakens in the case

of a manager not being able to engage in multiple bids prior to the current bid during his or her

professional career, hence some of them do not have a chance to learn from their experience,

notably young managers.

In addition to the extrapolation hypothesis and the self-attribution bias, many additional studies

have also explained causes and consequences of managerial overconfidence in the M&A

context. Moeller et al. (2004) affirm that bidder size plays an important role in managerial

overconfidence, suggesting that larger firms are likely to be overconfident as managers have

great discretionary power over the firms, resulting in fewer obstacles allocating a firm’s

resources. Since acquiring public firms is the quickest way of size expansion for a firm,

managers of large firms are likely to overpay for public targets with high growth opportunities.

Andreou et al. (2016) found a significant negative relationship between managerial


30

overconfidence and the value creation of corporate diversification. Analysing a sample of

diversification M&As, the authors found overconfident managers destroy firm value by 12.4

to 14.1 percent as opposed to rational managers do.

Despite the main stream of the literature that links managerial overconfidence and corporate

investments, which suggests that managers who overestimate their own ability make value-

destroying deals, a few papers note the bright side of managerial overconfidence. For example,

Galasso and Simcoe (2010) found a positive relation between managerial overconfidence and

a firm’s innovative performance. Using a CEO stock-option exercise for managerial

overconfidence consistent with that used by Malmendier and Tate (2005), the authors

discovered that overconfident managers tend to boost the innovative performance of a firm, as

overconfidence leads the managers to underestimate the probability of the failure of deals,

encouraging them to pursue innovative activities. In addition, overconfident managers are

likely to make significant changes in innovative strategies that make the firm stand out in a

competitive environment, which is more pronounced in competitive industries. Hirshleifer et

al. (2012) who measured managerial overconfidence with options and press-based proxies for

a sample of U.S. CEOs during 1993 to 2003 also found that overconfidence drives managers

to engage in riskier projects, greater investment in innovative activities and greater innovative

activities, suggesting that overconfident managers are better innovators and more likely to be

hired by high growth industries.

2.1.3.2. Investors’ overreactions and underreactions

Debondt and Thaler (1985) postulate the overreaction hypothesis that news-oriented investors

exhibit reactions driven by unexpected and dramatic news events, which lead to market

fluctuations and make it predictable. Using the monthly data of U.S. stock returns, the authors


31

uncovered evidence of inefficiency in a weak form market, thus proposing two testable

hypotheses. First, price reverses follow extreme price movements that have occurred over the

previous days. Second, such a price adjustment tends to be greater when the initial price

movement is dramatic. Their study reveals that portfolios that have lower returns in the past 5

years outperform those have higher returns by 25 percent, suggesting that the market is sluggish

to respond to fresh information and overreaction is due to mean-reversion.

Consistent with Debondt and Thaler (1985), Barberis and Vishny (1998) suggest that the

overreaction or underreaction is due to investors’ mistakes. When a company announces good

news, the market tends to be over-optimistic about the firm’s prospects, which pushes the

firm’s price to high levels. However, the price was revised when investors recognise the

fundamental reason for the price boost was due to overreactions. In a similar vein, investors

also revise their pessimism about loser stocks in the long term.

The overreaction hypothesis has been challenged by many researchers. Jordan and Pettengill

(1990), who duplicated Debondt and Thaler’s study (1985) in a different sample period, found

that losers become winners but winners do not lose. In addition, Fama and French (1996)

proposed a three-factor model to account for reversal of long-term returns as they argued size

effect on a firm’s abnormal returns, affirming that small firms are associated with higher risks,

while large firms are associated with lower risks. It became evident that small firms are likely

to generate higher returns than larger firms.

It is generally believed that evidence regarding the underration hypothesis is much robust than

that regarding the overreaction hypothesis. Barberis and Vishny (1998) documented the

mechanism for investors’ underreactions. When a company performs well (badly) the market


32

tends to raise (lower) its stock price and this will be further raised (lowered) when the

performance of the firm is well supported by its earnings performance in a subsequent period,

which often occurs in a short period of time (3 to 5 months) compared with the overreaction

(3 to 5 years). Whilst theorists who suggest high risks are associated with high returns, Barberis

and Vishny (1998) offered plausible explanations from a psychological perspective on how

investors make mistakes when processing new information, suggesting that managers can

formulate profitable investment strategies by taking full advantage of such mistakes. Edwards

(1968) suggests that conservatism bias of investors, suggesting that underreaction is formed

due to slow updating belief. In the light of this, Li and Yu (2012) found that conservative

investors focus on the market 52-week high and historical high, affirming the investors

attention hypothesis.

2.1.3.3. The misvaluation hypothesis

Shleifer and Vishny (2003) developed a model of M&As based on the assumption that

managers are rational and the market is less than fully rational. The model explains ‘who

acquires whom, the choice of the medium of payment, the valuation consequences of mergers

and merger waves’ (Shleifer and Vishny, 2003: 295). Their study focuses on the limitations of

the neo-classical theory of M&As, which focuses on industry-specific shocks, and the effect of

the method of payment on M&As, which is not clearly interpreted. In addition, reasons why

cash-financed M&As generate long-term positive earnings while stock-financed M&As

generate long-term negative earnings have not been well explained. So as to explore these

reasons, their study presents a simple valuation model of M&As to explain the short- and long-

term gains of the bidder. The model implies that the bidder can produce a positive synergy in

the short term yet few synergies in the long term. Important implications drawn from this model

are presented as follows:


33

Suppose the target firm is 0 and the bidder firm is 1. Stock volumes for firms 0 and 1 are K and

K1, and price per unit is Q and Q1, and Q1 being greater than Q. The price per unit for the

combined equity is S, q is the long-term price. Bidders would pay a price for the target, denoted

P. If synergies exist, P lies between Q and S (i.e. Q<P<S).

1. The combined market value is S (K+K1)-K1Q1-KQ, which is the value of the new combined

firms minus the value of each individual firms involved in an M&A deal.

2. The target value in the short term is (P-Q) K, suggesting that the offer price is over market

value.

3. The bidder value is made up of two parts, one part is the gains the bidders earned as long as

their pay is less than S, the other is the loss of value dilution of the bidder firm’s capital, from

Q1 to S or (S-P) K+(S-Q1) K.

4. Given that there is no long-term profit for the combined firm when acquisitions are funded

with cash since the gains to the target firm is the loss on the bidder firm, the effect on the value

of the target is K (P-q), whereas the value of the bidder is K (q-P).

5. For stock-financed acquisitions, suppose x is the ratio of deal value over combined firm

value or x = PK/S (K+K1). In the long term, this share is xq (K+K1) = q (P/S) K;

6. The net long-term gains to target shareholders in stock-financed acquisitions are q (P/S) K-

qK = qK (P/S-1), where qK is the stand-alone value of target in the long term, whereas for the

bidder shareholders, it is qK (1-P/S).


34

Shleifer and Vishny’s misvaluation model (2003) indicates that bidders are those with a

considerable valuation whereas targets are those with a low valuation. Further, the long-term

effect of cash-financed acquisitions suggests that the target is undervalued whereas the net

long-term effect of a stock deal generates a short-term loss and long-term gains, explaining

why an overvalued bidder tends to use stocks as ‘cheap currency’ to pay for targets. Moreover,

there is a reverse effect between the bidder firm and the target firm, i.e. what the bidder loses

is the target gains. Finally, the model reveals that both the target and the bidder firms could

gain in the short term if there is synergy from the mergers.

Shleifer and Vishny (2003) also reason that M&As are a trade-off game between the bidder

and the target, since they involve a high valuation bidder and a low valuation target based on

the rationale that a low valuation firm may find it hard to create synergy through the acquisition

of a high valuation firm. Their study suggests that a bidder buys a target to create long-term

value whereas the target focuses on short-term gains; hence, targets voluntarily accept

overvalued stocks for a cash-out purpose. All of these imply that bidder managers attempt to

maximise the interest of shareholders through exploring the market’s mispricing in an

inefficient market (Shleifer and Vishny, 2003: 296).

Therefore, Shleifer and Vishny (2003) assume that bidder managers are rational whereas target

managers are irrational as since they accept overvalued stocks. However, this view has been

challenged by Rhodes-Kropf and Vismanthan. (2004) who contend that rational managers

accept overvalued stocks by mistake. According to this study, target managers are misled when

the market valuation is high, and accept overvalued stocks as they overestimate synergies.

Further, Rhodes-Kropf et al. (2005) identified three types of specific valuation errors that drive

merger waves. By decomposing the conventional MTBV proxy into three components: the


35

firm-specific deviations from short-run industry pricing (firm-level misvaluation), sector-

factor, short-run deviations from a firm’s long-run mispricing (sector-level misvaluation), and

long-run pricing to book (long-run growth opportunities), the authors found that high MTBV

firms buy low MTBV firms, which is attributed to firm- and sector-level valuation errors. Firms

of low long-run growth opportunities tend to buy firms with high long-run growth opportunities.

Their study suggests a positive correlation between stock-financed acquisitions and market-

wide valuation.

In this connection, Dong et al. (2006) affirm that the misvaluation hypothesis is in favour of

high valuation periods. Analysing a sample of U.S. M&As from 1978 to 2000, the authors

claim that evidence for the misvaluation hypothesis is more relevant for M&As after 1990 than

before, whereas the evidence for the Q hypothesis is more relevant before 1990 than after.

Their findings suggest that the misvaluation hypothesis reconciles the case that the market-

wide valuation is high whereas the Q hypothesis is employed to explain how well-run bidders

explore synergies from poorly-performed targets. In addition, the authors find that more

overvalued bidders prefer to pay for targets with stocks rather than cash, choose mergers rather

than tender offers, and pay higher offer premiums. Their results are consistent with those

predicted by Shleifer and Vishny (2003).

Dong et al. (2006) investigated the effectiveness of the Q hypothesis and the misvaluation

hypothesis in different sample periods but fail to address the concern as to whether bidders

paying with overvalued stocks are being rational. According to Shleifer and Vishny (2003),

bidders who use overvalued stocks as a means of payment for M&As receive negative market

reaction in the short run, leading shareholders to believe that the firm does not generate

synergies. Their real purpose of using overvalued stocks is to dilute overvaluation and create


36

value for the firm, since holding stocks in an overvalued market incur risks for the firm (Jensen,

2005). In this respect, Ang and Cheng (2006) assessed the long-term performance of the bidder

who pays for acquisitions with stocks. Studying a sample of completed stock-financed

acquisitions, the authors find that bidders undertaking M&As with overvalued stocks perform

better than those engaging in M&As but not using overvalued stocks. Their study indicates that

not all M&As are driven by misvaluation, since their data show a small fraction of the stock-

financed acquisitions involve an undervalued bidder, which contrasts with Savour and Lu’s

view (2009) in their assumption that all stock-financed acquisitions are misvaluation-driven.

The misvaluation hypothesis has been challenged by many studies. Firstly, in line with Ang

and Cheng (2006), Fu et al. (2013) suggest that ‘The existence of relative overvaluation

between the bidder and the target is a necessary, but not a sufficient condition for the

acquisition to benefit acquirer shareholders’, as their M&A sample shows that two-thirds of

stock-financed acquisitions are motivated by a bidder firm’s overvaluation. They suggest that

bidder shareholders only benefit from overvalued stocks if low premiums are paid to the target

shareholders. Using a sample of 1,319 stock-financed acquisitions between 1985 and 2006, the

authors found that overvalued bidders do not outperform those firms in the control sample,

reflected in significantly worse operating performance and worse stock returns five years

following the M&As. By including corporate governance related proxies, their study reveals

severe agency problems in an overvalued firm, indicating that acquisitions driven by

misevaluation destroy a firm’s value when the firm experiences agency problems.

Further, Eckbo et al. (2016) suggest that ‘overvaluation reduces the all-stock payment

propensity’, which contradicts the idea of the misvaluation hypothesis. Using aggregate mutual

fund flows as an instrumental variable of the pricing error, the authors indicate that the shock


37

caused by mutual fund flows pushes up the market valuation but it does not increase the volume

of stock-financed acquisitions. Their study challenges the view that managers use overvalued

stocks to time the market. The authors find that bidders who pay with overvalued stocks for

financing acquisitions tend to be small firms and non-dividend paying firms with low leverage,

indicating that firms that are financially constrained avoid using cash as a means of payment

for financing M&As.

This section has reviewed three broad types of M&A motives. According to these studies

reviewed, there is no single theory that can fully address the M&A motive, since it varies

significantly in any particular sample at any particular time, which calls for up-to-date evidence.

Jensen and Ruback (1983), by reviewing a quantity of M&A research up to the 1980s,

suggesting that the primary M&A motive is synergies, and the small role the human factors

played on M&As weakened prior to 1990. This is also supported by a number of M&A research

works (Bradley et al., 1983; Bradley et al., 1988; Jensen, 1988; Lang et al., 1989; Servaes,

1991), whilst Jensen (1986) indicates the agency problem of free cash flow where managers of

cash-rich firms invest unwisely, and Roll (1986) suggests hubris-infected managers engage in

value-destroying M&A deals. Shleifer and Vishny (2003) suggest that M&As are driven by the

relative valuation of the two firms involved, indicating the role of managers’ timing the market.

Dong et al. (2006), who studied the misvaluation hypothesis and the Q hypothesis in the context

of M&As, noted that synergy-driven acquisitions are more likely to be found prior to 1990

whereas the misvaluation-driven acquisitions are common after 1990.

Chapter 2. Literature Review: M&A Process

38

2.2. M&A process

2.2.1. Selecting a target

Jemison and Sitkin (1986), who explored the M&A motive and M&A process, identified three

important M&A motives for bidder management: strategic fit, organisational fit and acquisition

process. The first two motives indicate that the acquisition should be financially or strategically

fit for the firm’s prospects, while the latter one indicates the essential role of M&A participants,

including analysts and key managers, played during the course of M&As. In this line of

thinking, human factors have a substantial impact on the M&A process.

The issue concerning “who buys whom” in the M&A context has been discussed for years.

Takeover targets are largely predicted within the scope of the bidders’ M&A motives. The

synergy hypothesis predicts that the less efficient firms are likely to be a takeover target, as the

bidder views target’s price discounts as a source of synergies generation. Investigating the

probability of takeover targets, Papelu (1986) proposed six hypotheses relating to “who buys

whom”, including the inefficient management hypothesis, the growth-resource mismatch

hypothesis, the industry disturbance hypothesis, the size hypothesis, the market-to-book

hypothesis and the price-earnings hypothesis, identifying that firms with an inefficient

management, smaller size, mismatch between resource and growth opportunity10 and high

leverage ratio are probably the target firm in an M&A deal.

Brar et al. (2009) who examined the probability of takeover targets by extending Palepu’s

model (1986) with inclusions of price momentum, trading volume and a measure of market

sentiment for a large sample of European and cross-border acquisitions found that takeover

10 According to Palepu (1986), the mismatch between resource and growth opportunity refers to a rich resource

with low growth opportunity or poor-resources with high growth opportunity.


39

targets are small, undervalued and less liquid relative to bidders. Their findings are in

accordance with Palepu (1986). In addition, the authors suggest that takeover targets are those

firms with low sales growth and strong short-term price momentum and their shares are

actively traded prior to the announcement date, indicating that arbitrage around the M&A

announcement date leads the price of the target to be hard to value. The authors also noted that

takeover targets are those firms with incompetent managers, inducing other managers to

explore the potential synergies through takeovers.

Jensen (1988) notes that managers who compete over the rights to control a firm’s resources

initiate M&As. Hence, M&As are a vehicle of integrating the bidder firms’ and the target firms’

resources, suggesting that poorly performing firms are likely to be acquired in M&As.

Similarly, Mitchel and Lehn (1990) found that firms whose managers made value-destroying

deals are likely to subsequently become takeover targets. Lang et al. (1989) and Servaes (1991)

also provided supportive evidence that inefficient firms are likely to be takeover targets. Their

findings suggest that acquisitions enhance value for the firm when they involve a high Q bidder

and a low Q target. The Q hypothesis indicates that high Q bidders are easier to explore

synergies from the poor-performing firms than from the well-performing firms through M&As.

Another line of literature suggests that it may not be the case that acquired firms are in a weak

position prior to an M&A. Bradly et al. (1983) documented that targets are those ‘sitting on the

gold mine’, suggesting that targets hold excess cash, hence, they are attractive to firms

experiencing financial constraints. Agrawal and Jeffe (2003) find direct evidence that targets

do not underperform prior to M&A announcements, and their results continue to hold

regardless of various measures employed to influence the firm’s performance. Rhodes-Kropf

and Robinson (2008) showed evidence of ‘like buys like’, suggesting M&As involve a firm


40

with similar prospects. Based on the relative bargaining power of the bidder and the target,

their model suggests that bidders tend to acquire firms of a similar MTBV. Levis (2011)

suggests that firms of high revenue growth opportunities but high operating costs are probable

takeover targets whereas firms of lower growth opportunities but high cost efficiency are

probable takeover bidders. Investigating a sample of newly-listed firms that become takeover

targets shortly after their initial public offerings, De and Jindra (2012) noted firms that perform

well in operations or in the stock market are likely to be acquired due to superior post-IPO

performance.

Studies have addressed the issue concerning “who buys whom” from other angles. Massa and

Zhang (2009) found M&As are driven by the popularity difference between the two firms

involved. They identified ‘cosmetic mergers’ where a less popular firm tends to buy a more

popular firm. This investment style is more prevalent when the popularity difference 11

between the target and the bidder is significant. The popularity of an acquired firm creates a

‘hole’ effect for a less popular firm, increasing the probability that the market will re-evaluate

the bidder firm. Shleifer and Vishny (2003), however, suggest that the undervalued firm is

probably the takeover target as it provides great scope for value dilution for the highly valued

firm. If the primary M&A motive of the bidder is to eliminate any takeover threats, larger firms

in the same industry as the bidders are the probable targets since they can quickly help expand

the bidder firm’s size and reinforce their competitiveness among the industry peers (Gorton et

al., 2009). Indeed, Hoberg and Phillip (2010) showed that firms offering the bidder some hard-

to-imitate assets are attractive takeover targets.

11 The authors used data on the flows in mutual funds in different styles to construct a measure of popularity.


41

Not only bidders look for targets, but also targets actively search for bidders (i.e. target-initiated

deals12). Studying a sample of over 3,800 U.S. public acquisitions between 1996 and 2009,

Eckbo et al. (2014) found 45% of acquisitions in their M&A sample were initiated by targets.

Their findings are also supported by Boone et al. (2007) who observed that about half of

takeovers of 400 U.S. takeovers announced between 1989 and 1999 had been initiated by

investment bankers hired by the target whose aim was to sell the firm. Masulis and Simsir

(2015) suggest a large number of M&As deals announced between 1997 and 2006 are target-

initiated: 237 as opposed to 408 bidder-initiated deals in their sample of 645 M&A deals. The

authors documented that target financial or economic weakness, target financial constraints and

economy-wide shocks were the key factors that determined target-initiated deals. The target-

initiated deals that are not strategically or financially fit for bidders reinforce bidder’s

bargaining power, resulting in lower levels of offer premiums, target announcement returns

and the deal value to EBITDA multiples of the target-initiated deals compared with the bidder-

initiated deals (Masulis and Simsir, 2015).

2.2.2. Paying for a target

Paying for a target is of first-order importance in M&As, since how much a bidder pays for

acquisitions directly determines the wealth distribution between the shareholders of the two

firms involved. Though a high offer premium increases the likelihood of the deal success, it is

at the expenses of the wealth of bidder shareholders. Because a low offer premium is hard to

persuade the target to give up the firm given that target shareholders are loss-averse (Baker et

al. 2012). However, the argument above leads the researcher to a question as to how much

offer premiums are too much, as on-one can give an accurate figure of the valuation of the firm

since valuing a target requires additional considerations regarding the decision makers. Two of

12 Information on who initiates a deal is not identified in SDC database.


42

the most important factors influencing M&A valuation is how much M&A offer premiums the

bidder pays for the target, and by what means the bidder is willing to pay. The following section

reviews literature on the offer premium and the method of payment.

2.2.2.1. M&A offer premiums

Schwert (1996) explains offer premiums as the sum of the target price run-up and the mark-

up.13 The author found that final offer premiums increase with the pre-bid target stock price

run-up while the price mark-up fails to explain a majority of offer premiums. His study

measured the target price run-up with the target abnormal stock returns from 42 days to 1 day

prior to the takeover announcement date relative to the abnormal stock returns during the

announcement date, and offer premiums with target abnormal stock returns between 42 days

prior to the takeover announcement date and 126 days after takeovers.14 Analysing a sample of

1,814 U.S. mergers and tender offers between 1975 and 1991, the author established that for a

one dollar increases in target price run-up would cost a bidder firm an extra 1.13 dollar.

Similarly, Betton et al. (2008), who studied the relationship between the target price run-up

and initial offer premiums, found that for a dollar increase in the price run-up increased the

offer price by 0.8 dollars.

The minimum offer price equals the current market price of the target firm given that the firm’s

value is fairly estimated. It is commonly believed that the offer price should be higher than the

market price of the target, otherwise the target will refuse the offer, arguing that it will bring

forth synergies to the combined firm. However, in some cases firms are sold below market

value or with negative offer premiums. Weitzel and Kling (2014) suggest that targets

13 In other words, offer premiums were defined as the difference between abnormal returns and the price run-up. 14 For those targets delisted before 126 days after takeovers, he used the day when they were delisted.


43

compensate bidders in the acquisitions on the rationale that they believe there will be post-

merger synergies. Secondly, targets compensate bidders who agree to accept their overvalued

shares, since small firms find it hard to justify overvaluation. Finally, targets who experience

severe shortage in liquidity tend to sell firms with negative offer premiums.

However, in most cases, the offer premium is positive and underlines the overpayment

hypothesis or the synergy hypothesis. If offer premiums are driven by wealth creation purpose,

bidder firms gain positive abnormal returns around the announcement date (i.e. the synergy

hypothesis). In agreement with this line of thinking, Antoniou et al. (2008) found that offer

premiums and bidder returns are positively related in a sample of 396 U.K. public acquisitions

announced between 1985 and 2004.

Evidence suggests lower offer premiums are paid to the target shareholders if the needs of

target managers are met. Wulf (2004) suggests bidders who guarantee the retention of target

managers in the combined firm tend to pay a lower offer premium for the target. Studying the

relationship between offer premiums and the choices of payment method in a sample of 2,959

acquisitions between 1983 and 2004, Zhang (2009) found that offer premiums decrease when

cash is paid as a result of increasing popularity of cash. Their findings suggest that cash

payments increase the bargaining power of the bidder when negotiating with targets who have

a strong cash preference. In this line of thinking, bidders, by offering what the target manager

wants, save takeover costs.

The overpayment hypothesis suggests that bidders tend to overbid for the target if the offer

premium is greater than expected synergies generated following the combined firm for a variety

of reasons. Roll (1986) indicates that bidders who overestimate the synergies of a combined


44

firm or over-optimistic about their own management skills tend to overpay for targets. Rhodes-

Kropf and Viswanathan (2004) explained this as overestimation about the synergies increasing

with valuation errors. Following this line of discussion in relevant literature, Hayward and

Hambrick (1997) provided related support that hubris-infected managers tend to overpay for

acquisitions. Investigating offer premiums for a sample of large U.S. M&A deals, the authors

suggest that CEOs whose firms have performed well recently, who are praised by the media or

have a high salary have strong sense of self-importance, which increase high offer premiums

to the target shareholders. Antoniou et al. (2008) found overpayment for diversified M&A

deals and attribute this phenomenon to over-optimistic CEOs stepping into an unrelated field.

Baker et al. (2012) found confident bidders who pay according to the target 52-week high so

as to successfully obtain an M&A deal, reasoning this as the reference point theory of an M&A.

Since targets have a loss-aversion tendency, a price that is significantly lower than this

reference point price will increase the probability of the deal being rejected by the target. The

authors also measured the market reactions to the M&A pricing decisions that depend on the

reference point, and found that the payment according to the target reference point was

associated with negative market reactions, as the market regards this as overpayment.

Rather than studies relating overpayment to bidders’ overconfidence or reference dependence

bias, others explaining it with the bidders’ rationality. Jensen (2005) suggests that bidders

overbidding the target with overvalued equity is a measure to avoid the risks of holding those

stocks in an overvalued market. The firm’s value will be destroyed when the initial

overvaluation is corrected by the market. This view is also supported by the prediction of

Shleifer and Vishny’s theoretical model (2003), suggesting that bidder managers attempting to

dilute their overvaluation are likely to overpay for a less overvalued target. In order to reduce

the risks of stocks being down-priced by the market, bidders offer high M&A premiums to


45

smooth the M&A process. Evidence that managers do not pay high offer premiums to the target

shareholders because of managerial irrationality is also supported by many additional studies

(Dong et al., 2006; Fu et al., 2013).

Studies have related overpayment to agency problems. Specifically, bidders’ overpayment is

due to lucrative compensation paid to managers who initiate or complete M&A deals. Harford

and Li (2007) and Guest (2009) suggest that CEOs engaging in an M&A deal are better off

regardless of the quality of the deal. Since the primary purpose for the manager is to obtain

rewards through M&As, they overpay for the target so as to smooth the M&A process. Jensen

(1986) also indicates that managers of cash-rich firms tend to pay excessively for these ill-

conceived investments.

It can therefore be concluded that offer premiums should bring forth positive abnormal returns

to the bidder if they are in line with the synergy hypothesis, while negative abnormal returns

will accrue to the bidder if they are explained with the overpayment hypothesis. In this pursuit,

Diaz et al. (2009) identified a quadratic relationship between the offer premium and bidder

abnormal returns, reasoning that the lower level of the offer premium represents its role of

synergies whereas the higher level represents overpayment.

2.2.2.2. Method of payment hypotheses

The choices of payment methods are another important factor influencing M&A valuation.

This section reviews four broad categories of explanations regarding the method of payment:

tax considerations, the signalling hypothesis, capital structure and corporate control and

behavioural finance, followed by relevant empirical evidence for each explanation.


46

2.2.2.2.1. Tax considerations

Studies have linked the method of payment with corporate taxes, suggesting that targets would

demand high offer premiums when M&As were solely financed by cash. In accordance with

capital gains taxes, there will be an immediate tax on capital gains for cash while tax on capital

gains for stocks is deferred until gains are realised. Therefore, high offer premiums are

associated with cash as a compensation for the targets. Wansley et al. (1983) who studied the

tax effect on the choice of payment method found that target cumulative average abnormal

returns (CAARs), measured with 41 trading day stock returns after M&As, are 33.54 percent

when the means of payment for M&As is cash, but this figure nearly doubles when it is

substituted with stock payments. Likewise, studying a sample of 204 pairs of mergers between

1977 and 1982, Huang and Walking (1987) reported CAARs for cash payments, stock

payments and the combination of cash and stocks were 29.3 percent, 14.4 percent and 23.3

percent respectively.

By contrast, Franks et al. (1988) found that the presence of the all-cash offer premium effect

was pronounced before capital gain taxes were introduced into the U.K. market, suggesting

little evidence of the tax treatment effect on the method of payment. They examined a sample

of 2,500 acquisitions in both the U.K. and the U.S. markets between 1955 and 1985 and found

that U.K. firms had a much stronger cash preference than their U.S. counterparts. In addition,

the period of 1965 to 1969 saw a significant decline in cash-financed acquisitions as compared

with that in the period of 1960 to1964, before the capital tax gains law came into effect: 18.6

percent compared with 29.2 percent. Moreover, their study indicates that cash payments,

though associated with higher offer premiums than stock payments, reap more benefits for

long-term shareholders due to the valuation effect.


47

It can be argued that early studies that examined the tax treatment effect on the choice of

payment methods based on the periods when the method of payment effect has received little

attention. In this line of thinking, examining the tax treatment effect on the method of payment

in a public U.S. acquisition sample between 1985 and 2004, Ismail and Krause (2010) provide

the most up-to-date evidence that fails to detect the tax treatment effect on the method of

payment. Their study reinforces Frank et al.’s findings (1988), suggesting the tax treatment

effect has little impact of on the method of payment.

2.2.2.2.2. Signalling hypothesis

Another classic explanation for the choice of payment methods is the signalling hypothesis

assuming that the market is transparent and there are no transaction costs and taxes, the method

of payment choices should be irrelevant to finance an M&A deal. However, Myers and Majuf

(1984) found evidence of information asymmetry, as managers know better than the market

about the value of a firm, in that they are able to choose the right sources of financing for

acquisitions. As such, stock payments signal to the market that the firm is overvalued and thus

incur negative market reactions, whilst a cash payment indicates that the firm is undervalued,

resulting in positive market reactions.

According to this literature, Travos (1987) found direct evidence that stock payments are

associated with negative bidder announcement returns whereas a cash payment is associated

with positive bidder announcement returns, claiming that stock payments convey bad news to

the market whereas as cash payment conveys good news to the market. Using a sample of

successful acquisitions between 1972 and 1982, the author examined a series of factors that

had a substantial impact on bidder announcement returns highlighted in the prior M&A

literature, including the method of payment, the offer premium, relative size and the type of


48

mergers. Surprisingly, it is suggested that only the method of payment has a significant impact

on bidder announcement returns, and stock payments are negatively related to bidder

announcement returns, which is consistent with M&A literature.

Hansen (1987) regards stock payments as a risk-sharing agreement between the bidder and

target firms. With high levels of information uncertainty, bidders are likely to be misled by the

valuation of the target. So as to avoid overpayment, bidders are prone to pay for acquisitions

with stocks, requiring the targets to share the downward risk of stock overvaluation, while cash

is a clear cut between the bidder and the target and paying acquisitions with cash indicate

undervalued bidders. His study underscores the central role of the risk relating to the method

of payment, providing a rationale on why bidders choose stocks for acquisitions given that

abnormal returns are lower than those paid with cash.

Fishman (1989) claims that cash payments played a pre-emptive role in M&As involving

multiple bidders. He proposed a model of pre-emptive bidding, suggesting that cash payment

signals to the target that the bidder’s valuation is high whereas stock payments indicate the

opposite case. The role that cash played in enhancing the advantages of the bidder was also

addressed in Franks et al.’s U.K. study (1988). Cornu and Isakov (2000) found supportive

evidence that cash offers deter competition, working better than debt or equity. Consistent with

the signalling hypothesis, Cornu and Isakov (2000) reason that cash offers signal to the market

a high-valuing bidder who is capable of managing the deal.

In the presence of information uncertainty, the market perceives stock payments as

overvaluation by the bidder or that uncertainty regarding expected synergies increases with

stock payments. As such, stock payments incur negative market reactions whereas cash


49

payment brings forth positive market reactions. However, a number of studies have challenged

this view, arguing that bidders do not underperform when financing an acquisition with stocks.

Savor and Lu (2009) found that bidders who are able to complete an M&A deal with stocks

would outperform those who fail to do so. Studying the joint effects of seasonal equity offerings

(SEOs) and method of payment on bidder gains, Golubov et al. (2015) found stock-financed

acquisitions are not the source of value-destruction of an M&A deal.

2.2.2.2.3. Capital structure and corporate control

The choices of payment methods also relate to a firm’s capital structure decision. Pecking order

theory suggests that firms prefer to use internal resources to fund investments and will not use

external resources until their internal resources are exhausted. The method of payment is the

means that the bidder chooses to pay for an acquisition, whereas the source of financing is what

the bidder funds an acquisition. The means of the bidder uses to get the funds can be different

from their payment decisions in M&As. Whilst conventional M&A studies considered the

means of payment as the same as the source of financing, Martynova and Renneboog (2009)

disentangled the effect of the source of takeover financing and the effect of the method of

payment. According to their study, the source of financing includes purely internal funds,

equity issues, debt issues and combinations of debt and equity issues. Acquisitions are paid

with cash when the source of financing is purely internal funds and debt issues, and are paid

with stocks when the sourcing of financing is equity issues, or a combination of equity and

debt and cash. Analysing a sample of 1,361 completed acquisitions in 26 European countries15,

the authors found that acquisitions funded by internal sources underperformed those funded by

debt issues.

15 Their sample consists of both European domestic and intra-European cross-border acquisitions with both

bidding firm and target firm from European countries or the U.K.


50

Pecking order theory is commonly violated in the context of M&As especially when bidders

are financially constrained or distressed. Martin (1996) noted that one of the main reasons that

a firm uses stocks as a means of payment for acquisitions is that the firm is in short of cash

prior to the M&A announcement date. Jensen (1986) found supportive evidence that a large

free cash flow motivates managers to use cash instead of stocks to finance an acquisition.

Harford et al. (2009) indicate that firms that have a higher leverage ratio relative to targets tend

to use equity as a means of payment for finance acquisitions while the lower leverage ratio

firms tend to use debt-financed cash. Their study relates the method of payment to the capital

structure of the firm. Firms in the face of larger risks of financial distress or bankruptcy are

more cautionary about the change in their leverage, indicating an important role of risks played

in the method of payment choices. In a similar vein, Alshwer et al. (2011) found financially

constrained bidders tend to retain internal sources to maintain financial flexibility.

Other studies have related the method of payment choices to managerial control. Managers

holding large stocks would have extensive control over the firm, in that they should be reluctant

to offer stocks in order to maintain their control. Stulz (1988) suggests that managers attemping

to protect their ownership of the firm avoid using stocks as a means of payment for acquisitions.

Martin (1996) obtained results consistent with of those in Stulz’s work (1988). The results

suggest a non-linear relationship between managerial ownership and the method of payment.

Managers are insensitive about the method of payment when their ownership is low, while a

high level of managerial ownership, between 5 and 15 percent, reduces the probability of

acquisitions being paid for with stocks, as managers with large stock holdings avoid using

stocks as this result in dilution of their control to over the firm. Similarly, Faccio and Masulis

(2005), who investigated a sample of European M&As between 1997 and 2000, noted that cash

is more likely to be used to fund acquisitions when the level of managerial ownership is


51

between 20 and 60 percent. These authors also document that the probability of stock payments

decreases when target firms have concentrated managerial ownership, since this creates new

block holders threatening the firm control of bidder managers.

However, it is argued that neither the target nor the bidder managerial ownerships have an

impact on the method of payment. Zhang (2001) examined the effect of managerial ownership

on the method of payment in the U.K. market using a number of factors that have a significant

impact on the method of payment, including the firm size, the financial policy of the bidder,

returns of equity and the performance of the bidder.16 Taking these controls into account, the

ownerships of the two firms involved showed little impact resulting from the method of

payment. The author concluded that there was little evidence of the managerial ownership

effect on the method of payment, but he failed to explain the rationale behind this finding.

2.2.2.2.4. Behavioural finance

Studies have emphasised the role of managerial behaviour played in the method of payment

choices. Shleifer and Vishny (2003) suggest that bidder managers choose an appropriate means

of payment for acquisitions. By doing so, bidder managers utilise stocks in acquisitions when

their firms are overvalued. Additionally, Vijh and Yang (2013) found small public targets were

less vulnerable to overvalued stocks. Studying a large sample of U.S. public acquisitions

between 1981 and 2004, the authors established a positive relationship between the target size

and stock-financed acquisitions, suggesting that small targets leave bidders limited room to

dilute overvaluation and acquiring small targets is associated with larger transaction costs and

higher offer premiums than large targets. This suggests that the valuation of firms is an

16 Zhang (2001) found that stocks were likely to be used for acquisitions when bidder firms were large and equity

returns were higher (as bidding firms might believe their stocks are undervalued therefore paying for acquisitions

with cash), while cash was likely to be employed for acquisitions when bidder firms paid higher dividends to their

shareholders and the bidder firm performed better in the stock market.


52

important factor for managerial considerations regarding the method of payment choices. Once

again, their evidence suggests rational managers link the method of payment to the valuation

of the firm.

Rhodes-Kropf and Visvanathan (2004) also claim misvaluation drives all stock-financed

acquisitions. Rhodes-Kropf et al. (2005) identified the three sources of valuation errors that

drive stock-financed acquisitions. Of these, the firm- and sector-level valuation errors were the

main explanatory factors. In support of this, Dong et al. (2006) found that stocks were likely

to be used for acquisitions as the two firms involved had a high valuation. They measured the

firms’ valuations with P/B and P/V, which are based on managerial perceptions, reasoning that

‘It could be argued that it is not actual misvaluation that influences takeovers, but merely an

incorrect perception by managers that there is overvaluation. If managers believe that the

price-to-book ratio is an indicator of misvaluation, they may make takeover decisions based

on this measure.’

Dong et al. (2006:756)

Following this, Ben-David et al. (2015) used short interest as a measure for mispricing where

short interest is a sophisticated investors’ belief regarding a firm’s misvaluation. The authors

suggest that bidders engaging in stock-financed acquisitions have stronger short interest, which

distinguishes the measure used for the real investment opportunities of the firm. The

mechanism is that short-sellers tend to take (avoid) a short position in overvalued (undervalued)

stocks. Further, their study suggested that firms in the top quintile of short interest are 54

percent more likely to engage in stock-financed acquisitions, which is consistent with the

misvaluation hypothesis.


53

Studying the willingness of targets to accept overvalued stocks, Burch et al. (2012) found that

institutions tend to accept or hold stocks with high valuation. It was found more than 56 percent

of the institutions in their sample sell target stocks before the completion of an M&A or convert

them to bidder stocks soon after that. Contradicting the view that targets accept overvalued

stocks by mistake (Rhodes-Kropf and Visvanathan, 2004)17, Burch et al. (2012) found that

institutional shareholders time the market using highly valued stocks. Managers find it easier

to use overvalued stocks to liquidate their holding position. They can sell stocks before the

market downgrades the value of stocks. In addition, they presumably believe that highly valued

firms are also well-performing firms or those with better investment opportunities. Hence,

holding stocks in these firms can generate benefits in the future.

Studies that examine the misvaluation hypothesis largely borrow MTBV as a measure for

mispricing, (Rhodes-Kropf and Visvanathan, 2005; Dong et al., 2006). However, Di Giuli

(2013) argued that MTBV is problematic for misvaluation as it is a proxy for both mispricing

and investment opportunities. Specifically, a firm that has better investment opportunity or is

highly valued leads to the same prediction that its MTBV is high. This is even misleading in

industries of high growth prospects. So as to disentangle the two effects, the author proposed

a new measure for investment opportunities, which is the average ratio of capital expenditures

over assets in the four years following the merger. The proxy is based on post-merger

investments. Using this on a sample of 1,187 mergers between 1990 and 2005, the author found

that one standard deviation increase in capital expenditure increases the probability of paying

acquisitions with stocks rather than cash by 7.5 percent, suggesting that better investment

opportunities lead bidders to increase stock payments for acquisitions.

17 The mistake refers to an overestimation of the synergy or underestimation of the market-wide overvaluation.


54

Rather than the misvaluation hypothesis, which predicts cash as a means of payment for

undervalued targets, the supply-driven explanation proposed by Zhang (2009) suggests that

cash is likely to be used as a means of payment for acquisitions when the popularity of cash

for target shareholders is high and target firms are less active and more constrained. Offering

what the target shareholders want, bidders also reinforce their bargaining position in an M&A

deal. It was found that cash paid to target shareholders with greater cash popularity is associated

with lower offer premiums.

Additional factors also exhibit a significant impact on the method of payment include. First,

the size of the firm relates to the method of payment. Dong et al. (2006) suggest that equity is

more likely to be used for acquisitions involving large bidder and target firms. It can be

interpreted as follows: bidders may find it hard to borrow large amount of cash to finance a

large target and thus use stocks as a substitute (Martin, 1996; Moeller et al. 2004). Further,

tender offers are more likely to be associated with cash payment whereas mergers are more

likely to be associated with stock payments (Martin, 1996). In addition, cash is likely to finance

hostile takeovers and generate higher abnormal returns for bidders (Schwert, 2000). In a cross-

border M&A settlement, the legal environment differences across the countries also affect the

method of payment (Martynova and Renneboog, 2009).

Chapter 2. Literature Review: M&A Performance

55

2.3. M&A performance

M&A performance is the consequence of M&As, which is one of the direct criteria that assess

the success of an M&A deal. It is a well-documented fact that targets receive positive market

reactions around the M&A announcement date, whereas bidders on average do not gain (Dodd

and Ruback, 1977; Jensen and Ruback, 1983; Rau and Vermaelen, 1998; Fuller et al., 2002;

Moeller et al., 2004; Alexandridis et al., 2013). As abnormal returns for short-term period may

not fully capture the announcement effect of a takeover, studies have also assessed the long-

term performance of the firm (Loughran and Vijh, 1997; Rau and Vermaelen, 1998; Agrawal

and Jaffe, 2000), since it reflects the holding experience of the investors. Some studies have

also assessed the operating performance of the firm in the post-merger period, which is an

accounting-based approach (Healy et al., 1992; Ghosh, 2001; Andrade et al., 2001; Martynova

et al., 2006). Overall, M&A performance is sensitive to the sample period examined18, the

sample criteria selected and the benchmark models employed to calculate abnormal returns of

a firm.

A number of factors have had a significant impact on the abnormal returns of the firm, and

been summarised into two broad categories: firm-specific characteristics and deal-specific

characteristics. Studies that link the M&A performance of firms include those variables as

standard controls, such as MTBV (Rau and Vermaelen, 1998; Sudarsanam and Mahate, 2003),

the size of the firm (Asquith et al., 1983; Moeller et al., 2004; Alexandridis et al., 2013), the

method of payment (Travos, 1987; Franks et al., 1988; Shleifer and Vishny, 2003), target status

(Fuller et al., 2002; Antoniou et al., 2007), merger choices (Dodd and Ruback, 1977;

18 There are six merger waves defined in the M&A literature (Martynova and Renneboog, 2008). The first merger

wave was between 1890 and 1903, the second merger wave was between 1910 and 1929, the third merger wave

was between 1950 and 1973, the fourth wave is between 1981 and 1989, the fifth wave was between 1993 and

2001, with the sixth merger starting in 2003.


56

Berkovitch and Khanna, 1991), deal attitude (Schwert, 2000; Martynova and Renneboog, 2006)

and industry relatedness (Morck et al., 1990; Agrawal et al., 1992).

2.3.1. Short-term performance

Numerous studies have examined the abnormal returns for the bidder firm, target firm, and the

combined firm with the event study. This approach has become dominated in analysing the

abnormal returns of the firms since it was first adopted by Fama et al. (1969). According to

Jensen and Ruback (1983), abnormal returns are the difference between the realised returns of

the firm and the benchmark returns:

‘Abnormal returns are measured by the difference between actual and expected stock returns.

The expected stock return is measured conditional on a realised return on a market index to

take account of the influence of market-wide events on the returns of individual securities.’

Jensen and Ruback (1983:6)

Early event studies measure abnormal returns of the firm around the effective date of merger.

However, Dodd and Ruback (1977) argued that the price effect could be released prior to the

announcement date, leading the price effect surrounding the completion date to be insignificant

(Asquith, 1983). Martynova and Renneboog (2011) proposed the necessity of observing both

the pre-announcement and post-announcement returns in an M&A study, since both are able

to measure M&A performance of the firm. Specifically, pre-announcement returns could

capture the information leakage, insider trading and market anticipation of an M&A deal while

post-announcement returns could reflect market correction for investor sentiment, such that

they over- or under-react to the M&A deal.


57

Martynova and Renneboog (2011) hold that M&A information is partially released prior to the

M&A announcement date, suggesting that investors could learn about the bidder’s objective of

takeover, the target’s attitude towards to the bid and so forth prior to the announcement date.

As such, abnormal returns surrounding the announcement date of a merger may not be able to

fully reflect the market reactions to an M&A bid. Following this view, Schwert (1996)

measured investors’ reaction to the M&A announcement with stock price change in a period

of 42 trading days prior to the announcement date and the announcement date. Stock price

change in the post-announcement period captures additional information that pre-

announcement period fails to do, such as, the level of post-merger integration and the

performance of successful and failed deals. However, it is difficult to assess the post-

announcement effect separate from other effects that also lead to a change of stock price

following a merger, such as financial or operational performance (Martynova and Renneboog,

2008). In spite of these factors, studies that examine the wealth creation from M&As have

employed a relatively short event window holding that price change around the announcement

date fully reflects M&A performance (Asquith, 1983; Dodd, 1980).

M&A studies have revealed appealing features of abnormal returns of firms by merger waves.

It is well-established that M&As generally create value for the two firms involved prior to the

fifth merger wave starting in 1993. Investigating the price change of the entire merger process

between 1962 and 1976, Asquith (1983) found positive announcement returns for bidder and

target firms regardless of deal outcomes. More specifically, targets enjoyed 6.2 percent two-

day excess returns19 for successful deals and this figure up to 7 percent for unsuccessful deals

19 The two-day excess return is the sum of abnormal returns in the window one day prior to the takeover

announcement date and the takeover announcement date.


58

while bidders enjoyed 0.2 percent two-day excess returns for successful deals and 0.5 percent

for unsuccessful deals.

Reviewing thirteen studies examining abnormal returns of both bidders and targets between

1950s and 1980s, Jensen and Ruback (1983) found that targets gain from M&As, whereas

bidders do not loss on average in abnormal returns around the announcement date. Of particular

note, targets outperform bidders in the samples of successful and unsuccessful M&A deals,

whereas bidders generally gain in successful M&A deals. In addition, bidder abnormal returns

are high regardless of whether the merger choice is a tender offer or merger for successful

M&A deals, and the two firms suffer small losses in abnormal returns for failed tender offers

and mergers. Franks and Harris (1989), who conducted an out-of-sample test, reported target

announcement returns in the range between 25 and 30 percent while bidders earn zero or

modest gains based on a sample of 1,800 U.K. takeovers announced in the period of 1955 to

1985, consistent with those in the work of Jensen and Ruback (1983).

Studies have shown significantly positive bidder announcement returns prior to 1993.

Analysing a sample of U.S. tender offers between 1958 and 1978, Dodd and Ruback (1977)

found that bidders earned 2.83 percent abnormal returns in a period of the announcement date

until 20 days following a takeover. Focusing on a similar sample period of Dodd and Ruback

(1977), Franks et al. (1977) reported abnormal returns of 4.6 percent for U.K. bidders. Studying

a comprehensive U.S. M&A sample, Schwert (1996) found both the two firms earn

significantly positive announcement returns regardless of merger choice. Eckbo and Thorburn

(2000), focusing on a sample consisting of both domestic and cross-border M&As between

1964 and 1983, found that abnormal returns to a U.S. bidder’s acquisition of a Canadian firm

were zero, whereas a Canadian bidder earned significantly positive in domestic M&As. Jensen


59

(1988) explained positive bidder abnormal returns as the effectiveness of the market for

corporate control returns prior to 1990s when managers compete to improve the efficiency of

the firm through M&As.

Whilst studies have found significantly negative bidder abnormal returns around the

announcement date after the start of the fifth merger wave, identifying various sources of value

destruction, such as agency problems, managerial overconfidence and market mispricing.

Agency problems suggest that the value of the firm is likely to be transferred from a firm that

is associated with severe conflicts of interests between the shareholder and the manager.

Harford et al. (2012) found entrenched bidder managers engaging in a value-destroying M&A

for fear of their stocks being diluted avoid using all equity, which in return destroys the firm’s

value. Moreover, entrenched managers avoid biding private targets that can potentially create

value for the combined firm, and they overpay for public firms and firms of low synergies.

Doukas and Petmezas (2007), who examined the effect of managerial overconfidence on bidder

gains between 1980 and 2004, found that bidder announcement returns were lower for a higher

order acquisition than a lower order acquisition, explaining this as the managerial self-

attribution bias. Using a large sample of U.S. mergers between 1978 and 2000, Dong et al.

(2006) found negative abnormal returns for overvalued bidders’ acquisition of a less

overvalued target. In such a case, bidder managers sacrifice the wealth of shareholders in the

short term to protect the wealth of the shareholders in the long term, and this finding is also

supported by Ang and Cheng (2006).

The sixth merger wave that started in 2003 witnessed some similar patterns of M&A

performance as those in the fifth merger wave. Alexandridis et al. (2010) suggest that bidder

abnormal returns decrease with the competitiveness of the takeover market. Studying a


60

worldwide M&A sample consisting of M&As announced between 1990 and 2007, the authors

suggest that bidders in the takeover competitive markets (i.e. U.S., U.K. and Canada) tend to

overpay, resulting in loss in the announcement returns, whereas bidders beyond these three

countries on average realise gains. Their study suggests that bidder abnormal returns vary

significantly from market to market. In spite of lower offer premiums are paid for targets during

the sixth merger wave than the fifth merger wave, bidders do not gain. Alexandridis et al. (2012)

underlined this to free cash flow problem, suggesting that the firms undertaking M&As in the

sixth merger wave tend to have lower valuation and higher cash balances than those in the fifth

merger wave, indicating that cash payment dominates stock payments in this period.

2.3.2. Long-term performance

2.3.2.1. Stock performance

A large body of M&A research has documented significant negative bidder abnormal returns

in the long-term up to a five-year period following a merger (Asquith, 1983; Jensen and Rubuck,

1983; Agrawal et al., 1992; Loughran and Vijh, 1997; Rau and Vermaelen, 1998; Franks and

Harris, 1989) 20. Asquith (1983), examining the bidder’s performance in the post-acquisition

up to 240 trading days following a takeover announcement date, reported a dramatic decline in

bidder abnormal returns compared with those on the outcome date of a merger. Specifically,

bidder abnormal returns decreased by 1.9 percent in the 100 days and by 7.2 percent in the 240

days following the date of the deal outcome. Jensen and Ruback (1983) reported negative

bidder abnormal returns of -5.5 percent in the year following a merger and noted that the result

is unsettling as it indicates market inefficiency. They further suggested this finding as the

market’s overestimation of the synergies accompanying a merger. Ruback (1988) late accepted

20 Despite these studies, Higson and Elliott (1998) reported zero abnormal returns in the three years following a

merger to the U.K. bidders between 1975 and 1990. In addition, equal-weighted abnormal returns to bidders were

found to be insignificant while bidders whose abnormal returns were calculated with the value-weighted returns

were positive and small.


61

the result that significant negative abnormal returns to bidders over the two years following a

merger as a fact. The notion that the market is inefficient has been supported by many

additional studies in favour of negative bidder long-term abnormal returns. Agrawal et al.

(1992), among other researchers, also reported negative bidder abnormal returns of -10 percent

over five years after a merger and their results continue to hold after controlling for size effect

and beta risk in various benchmark models.

Agrawal and Jaffe (2000), citing the work of Agrawal et al. (1992), attribute underperformance

as a ‘post-merger performance puzzle’. The authors outlined the method of payment and the

performance extrapolation hypotheses to this underperformance puzzle. The former hypothesis

indicates that when equity is used, the bidder is likely to be overvalued, and when cash is used,

the bidder is undervalued. Since the overvaluation will be corrected by the market in the long

term, stock-financed acquisitions are associated with negative market reactions, while cash-

financed acquisitions generate positive market reactions. The performance extrapolation

hypothesis indicates that the market overestimates the firms with high MTBV, which creates a

strong momentum around the takeover announcement date and leads to a strong long-term

reversal. Thus, Agrawal and Jaffe’s findings (2000) are consistent with those of Rau and

Vermaelen (1998). In a similar vein, Gregory (1997), who examined the relationship between

long-term abnormal returns and the free cash flow of a firm in the U.K. market, noted that firms

with high Q and low free cash flow are overvalued and underperform in the market in the long

term.

Another explanation for bidders’ underperformance in the long term is the neglect of the post-

merger integration process. Hansen (1987) indicates larger potential revaluation loss is

associated with the greater relative size of the target to the bidder. Moeller et al. (2004) suggest


62

that large deals require more diligent work by bidders, and they are associated with

considerable integration costs that lead to downward synergies. Despite their study not directly

addressing the long-term performance of the bidders, managers of a larger firm are likely to be

hubris-infected due to extensive managerial autonomy in the firm, causing more negative

announcement returns. Likewise, Antoniou et al. (2008) re-addressed the loss in the long-term

abnormal returns as the neglect of post-merger integration in a sample of U.K. public

acquisitions.

Others attribute bidders’ underperformance in the post-merger period to misspecifications of

the benchmark models employed. Franks et al. (1991), who tested a sample of 399 U.S.

takeovers between 1975 and 1984, suggested underperformance is measurement errors of the

models. The authors use multiple benchmark models in assessing the long-term performance

of the bidder and find insignificant abnormal returns, and the results are significantly different

depending on which portfolio is used: the equally- or value-weighted. Moreover, the authors

claim that long-term abnormal returns are largely dependent upon the relative size of the target

and the bidder, the method of payment, and the models employed. Barber and Lyon (1997)

also addressed bad model problems. They proposed multifactor models instead of a single

model to avoid mean-variance inefficiencies, and control firms of similar size and MTBV to

calculate long-term abnormal returns within five years following a merger. Their approaches

showed well-specified test statistics.

2.3.2.2. Operating performance

Researchers also measure bidder long-term performance using operating performance by

assessing the firm’s performance from its fundamentals, avoiding measurement errors from the

benchmark models based on an efficiency market. (Healy et al., 1992; Barber and Lyon, 1996;


63

Ghosh, 2001; Andrade et al., 2001; Martynova et al., 2006). According to Powell and Stark

(2005), previous M&A studies for assessing a firm’s operating performance following a merger

were prepared from three perspectives: the measures, the deflator choices21 and benchmark

models. Barber and Lyon (1996) suggest the important roles that firm size and pre-

announcement performance played in operating performance, showing test statistics well when

controlling for these two factors in the matched firm sample.

Studies commonly employ cash flows as a proxy to assess a firm’s operating performance

following a merger (Healy et al., 1992; Barber and Lyon, 1996), as other accounting-based

measures are likely to be manipulated after takeovers (Powell and Stark, 2005). Healy et al.

(1992), examining a sample of 50 largest U.S. publicly industrial mergers completed between

1979 and mid-1984 and using operating cash flows and asset productivity as proxies for

operating performance, found strong operating improvements following a merger. In a further

analysis, the authors found a positive relationship between abnormal returns and operating cash

flow performance of the bidder, implying that the market can predict the operating performance

of the firm.

Likewise, Switzer (1996) in his study focusing on a relative large sample size finds firms’

operating performance improved following a merger, regardless of the firm’s size, relatedness

of the two firms and bidder leverage. Linn and Switzer (2001) also provide evidence in support

of the existence of improvement in a post-merger performance. According to their study, cash

payment is associated with significantly greater change in operating performance than stock

payments. Their results indicate that bidders use cash as they have private information about

21 Powell and Stark (2005) have reviewed a series of studies relating to the techniques used in scaling the

accounting measures.


64

the potential synergies to be realised in the long term, reflected in an improvement of operating

performance.

However, Ghosh (2001) challenges the findings of Healy et al. (1992) by arguing that such an

improvement in operating performance is due to the fact that firms engaging in acquisitions are

larger firms compared with their industry peers. In most cases, large firms enjoy superior

performance, thus examining the operating performance of the firm exists bias when including

large firms. To combat this concern, the author uses firms matched on size and performance as

a benchmark to account for pre-performance of the bidder and compares the operating

performance of the firm before and after mergers with the matched firm sample. His study

shows little evidence that operating cash flow performance improves following a merger.

Many additional studies borrowing different operating measures have shown a decline to

operating performance following a merger (Mueller, 1980; Clark and Ofek, 1994; Fu et al.,

201322). Mueller (1980) observed a decrease of operating performance measured with ROE,

sales growth rate and total asset growth rate. Clark and Ofek (1994), studying a sample of

M&As with financially distressed targets and using earnings before interests, taxes and

depreciation (EBITD) as a proxy23 for operating performance also document a decline in

operating performance.

Using different accounting-based measures for the firm’s operating performance, Gugler et al.

(2003) reported mixed results in a sample of worldwide M&As announced between 1981 and

1998. More specifically, measuring operating performance with Profit/Assets, the authors were

22 Fu et al. (2013) detected the operating performance by focusing on a test for an implication of the misvaluation

hypothesis of Shleifer and Vishny (2003), which is whether the long-term performance of the stock bidders

improves when they use overvalued stocks as a means of payment for acquisitions. 23 His study scaled EBITD with revenues, which is EBITD/Revenues.


65

able to show an improvement in operating performance while the results decreased when

measured with Sales/Assets. Similarly, Martynova and Renneboog (2006) used four different

accounting-based measures to detect the operating performance for a sample of European and

U.K. M&As between 1997 and 2001. Controlling for industry, size and pre-takeover

performance in the matched firm sample, the authors showed that the previous significant

decrease in operating performance tends to be insignificant. In addition, the authors indicate

when the intercept model (or the regression-based model) was applied, the profitability tended

to be significantly positive and when the change model used the profitability tends to decrease.

Once again, their studies indicated that a firm’s long-term operating performance was sensitive

to the factors controlled and benchmark models employed.

2.3.3. Factors influencing M&A performance

2.3.3.1. The method of payment

The performance of the firms involved in an M&A deal are dependent upon a number of factors

examined in M&A literature (Datta et al., 1992; King et al., 2004; Eckbo, 2009), and the

method of payment has been widely studied in this regard. A review of the method of payment

hypotheses was presented in the 2.2.4 section of this thesis.

There is unanimous agreement among researchers that cash payment outperforms stock

payments in the short term. Studying a sample of U.S. acquisitions in the period of 1980 to

2001, Moeller et al. (2004) reported 1.38 percent bidder abnormal returns for acquisitions were

financed by cash, which is higher than stocks (0.15%). Martynova and Renneboog (2006), who

employed an 11-day window to calculate the bidder cumulative abnormal returns in a sample

of E.U. acquisitions, also reported quite similar results. In their study, though stock-financed


66

acquisitions can generate positive abnormal returns to bidders, they are significantly lower than

the returns on cash payment, 11.1 percent as opposed to 20.17 percent.

Similar results were reported in the M&A studies that focus on the bidders’ long-term abnormal

returns. Loughran and Vijh (1997) examining 947 acquisitions in the period 1970 to 1989 noted

that stock-financed mergers significantly underperformed cash-financed tender offers in the

five years following a merger, -25 percent as opposed to 65 percent. These authors provided

two proposed reasons in support of this finding. First, it was found that M&As that are paid in

cash are also tender offers that replace inefficient management and thus improve a firm’s value.

Secondly, bidders pay with stocks if the stocks are overvalued, leading to a decline in value

after market correction. In this vein, Bouwman et al. (2003), examining a sample of U.S.

acquisitions between 1979 and 1998 and using a two-year period following the acquisition,

reported negative excess returns of 7.03 percent for all stock-financed mergers whereas there

were insignificant excess returns of -1.76 percent for their cash counterparts. Analysing bidders’

long-term abnormal returns in a three-year period following an acquisition in a U.S. sample

between 1984 and 2001, Ang and Cheng (2006) reported negative excess returns of -12.45

percent for all-stock financed acquisitions, the result the authors explained with the

misvaluation hypothesis. Sudarsanam and Mahate (2003), focusing the U.K. market, reported

significantly negative abnormal returns for stock-financed acquisitions in the period 2 to 36

months following an acquisition regardless of the level of the price to bidders’ earnings ratio.

Researchers have addressed the effect of the payment method on the firm’s performance by

different M&A subsamples. Bouwman et al. (2003), who examined bidder abnormal returns

by merger choice, indicated that abnormal returns to bidders were lower for acquisitions

financed by stocks compared with those financed by cash, the result is more significant for the


67

merger subsample than the tender offer subsample. In a sample of U.S. acquisitions between

1979 and 1998 and to calculate bidder abnormal returns with a 3-day window, the authors

demonstrated insignificant positive abnormal returns for a sample of tender offers financed by

cash and stocks: 0.36 percent and -0.62 percent, while for those mergers, bidders enjoy

significantly positive abnormal returns for cash-financed acquisitions (0.88%) and significantly

negative abnormal returns for stock-financed acquisitions (-0.79%).

Whilst numerous M&A studies documenting evidence show that cash payment generates

positive abnormal returns and stock payments generate negative abnormal returns for bidders,

Chang (1998) reported significantly negative bidder abnormal returns for stock-financed public

acquisitions (-2.46%), but insignificantly negative bidder abnormal returns for cash-financed

public acquisitions (-0.02%). However, these results are reversed when targets are private firms

(i.e. unlisted firms). In particular, bidder abnormal returns for stock-financed private

acquisitions are significantly positive (2.64%) whereas cash-financed private acquisitions are

insignificantly positive (0.09%). This author provided three hypotheses to underline this

finding: the limited competition hypothesis, the monitoring hypothesis and the information

hypothesis. The limited competition hypothesis indicates that positive bidder abnormal returns

are due to the high probability of underpayment given weak competition for private

acquisitions. The monitoring hypothesis indicates outstanding performance achieved by

bidders using stocks for private acquisitions, as paying with stocks creates new block holders

that motivate the firm’s performance. The information asymmetry hypothesis indicates that the

market may re-evaluate the bidder’s valuation and reacts positively to the bid announcement

based on the rationale that private targets value stocks with caution. This finding is also

supported by Draper and Paudyal (2006) who have examined the U.K. takeover market where

a large majority of deals involve a private target firm.


68

2.3.3.2. Target status

Target status plays an important role in explaining bidder announcement returns. In most cases,

a private target is an illiquid firm, motivating the bidder to restore price discounts through

takeovers. Furthermore, a private target has a weak bargaining position, which reduces M&A

offer premiums. Studying a sample of 3,135 acquisitions between 1990 and 2000 and using a

5-day window for CARs, Fuller et al. (2002) found significantly positive returns were created

for bidder firms. When looking into subsamples according to target status, the authors noted

that bidder announcement returns were -1.0 percent when the target was a publicly-listed firm,

2.1 percent when the target was a privately-held firm and 2.8 percent when the target was a

subsidiary. Results are similar to Moeller et al.’s work (2007) employing a 3-day window for

CARs. Moller et al. (2007) studied a sample of 4,322 U.S. acquisitions announced between

1980 and 2002 and found that for all stock-financed acquisitions, the value of the bidder

declined by 2.3 percent when the target was a public firm, whereas it increased by 3.4 percent

when the target was a private firm. Their results are also supported by Conn et al. (2005) who

reported significantly loss in bidder abnormal returns in acquisitions involving a public target

when analysing a sample of U.K. M&As between 1984 and 2000.

Contradicting this line of studies, Bradley and Sundaram (2004) show that private firms

decrease the wealth effect of the bidder. Analysing a sample of over 12,000 U.S. acquisitions

between 1990 and 2000 and measuring abnormal returns in a window of 1 day and 24 days

after the M&A announcement date, the authors found that bidder announcement returns were

significantly negative (-10.09%). In particular, all-stock payments decreased the value of the

bidder firm by 6.35 percent for public acquisitions, whereas a decrease of 14 percent was noted

for private acquisitions. It can be argued that private firms are associated with high risks as


69

their information is less transparent than public firms, hampering synergy estimation of the

target firm.

2.3.3.3. Merger relatedness

Unrelated acquisitions24 refer to the two firms involved in an M&A deal being from different

industries. Researchers have not reached a consensus on whether diversifying acquisitions

improves the value of a firm. Datta et al. (1992) in a review study identified the wealth of

shareholders in a related acquisition was created through a transfer of core skills, economies of

scale, economies of scope and the value enhancement from the market power. The wealth of

shareholders in an unrelated acquisition is created through risk diversification, cheaper access

to capital, and value enhancement resulting from operational efficiency.

Morck et al. (1990) found diversified acquisitions are associated with lower bidder abnormal

returns, reasoning that there was a lack of experience when managers step into an unrelated

field. In support of this, Berger and Ofek (1995), by comparing the stand-alone value of the

firm and the market value of the firm in a diversified M&A deal, find that the market value of

the firm following a diversified acquisition was on average between 13 percent and 15 percent

loss in abnormal returns. Denis et al. (2002) claimed that neither global diversification strategy

nor industrial diversification strategy can create value for the firm. Contradicting this thread of

literature, Comment and Jarrell (1995) noted that corporate diversification increased the wealth

of shareholders in 1980s, identifying that diversification strategy as enhancing the firm’s value

through economies of scale.

24 Also known as diversified or conglomerate acquisitions.


70

2.3.3.4. Merger attitudes

The attitude of a merger can be broadly split into two, namely, solicited (friendly) and

unsolicited (hostile). A hostile deal refers to a deal opposed by the board of the target

management. Servaes (1991) reports loss in bidder announcement returns of hostile deals,

reasoning that the hostile bid results in high offer premiums or an increase of the probability

that a deal will fail. And in either case, bidders receive negative market reactions. In contrast,

friendly bids create value for both the bidder and the target firms, according to Morck et al.

(1988).

Schwert (2000) wrote that “the distinction between hostile and friendly offers is largely a

reflection of negotiation strategy” and “hostile takeovers mean different things to different

people”, implying that hostile takeovers do not reveal useful information to the market. The

author finds little evidence of the relationship between bidder announcement returns and hostile

deals based on the SDC data. The findings support the view that target managers opposing

hostile bids either because they are entrenched or because they use it as a bargaining strategy,

which contradicts Lang et al.’s view (1989) that target managers opposing hostile takeovers

signals to the market that their firm is well managed and creates value for the combined firm.

2.3.3.5. Merger choice

The choice of an acquisition includes tender offers and mergers. Tender offers are a share

purchase directly proposed to target shareholders, whereas mergers involve a negotiation

between the boards of the two firms involved. There are two advantages to initiating an

acquisition with tender offers, rather than mergers. First, tender offers avoid lengthy

negotiation processes. Golubov et al. (2012) indicate that the speed of deal completion for

tender offer is greater than that of the merger. They found that tender offers are approximately


71

55 days shorter than mergers in a sample of completed bids: 1,650 completed deals out of a

1,836 M&A sample. Secondly, initiating tender offers does not need the prior approval or a

contract with the target management (Betton et al., 2008), reducing bidder managers’ efforts.

However, tender offers are relatively costly for bidders. Offenberg and Pirinsky (2015), by

analysing a sample of hand-collected U.S. acquisitions announced between 2007 and 2012,

found that tender offers were associated with higher offer premiums than mergers since tender

offers may indicate the market that targets would generate large synergies for the combined

firm. In this sense, the authors should expect a positive relationship between tender offers and

bidder gains. However, they do not provide direct evidence in support of this implication.

Instead, they find that tender offers incur negative bidder abnormal returns, as bidders initiating

a tender offer are regarded as bad news. Moeller et al. (2004) support their view with positive

bidder abnormal returns associated with tender in a sample of completed U.S. acquisitions25.

The finding is more appealing for tender offers involving a larger bidder than a smaller bidder

(Offenberg and Pirinsky, 2015). It can be concluded that tender offers, though costly, reduce

the speed of the M&A process that turns out to be effective for M&As.

It is expected the use of tender offers increase the firms’ M&A performance. Jensen and

Ruback (1983) indicate that tender offers are associated with higher bidder and target

announcement returns than those of mergers. They reported 29 percent of abnormal returns to

target shareholders in tender offers compared with 16 percent to mergers, and 4 percent in

tender offers compared with zero gains in mergers. Franks and Harris (1989) reported 24

percent abnormal returns in the month of the announcement date to the target shareholders,

25 Offenberg and Pirinsky (2015) suggest that tender offers are more likely to be associated the deal is completed

than mergers.


72

which are significantly higher than those recorded in tender offers (14.8%). Bidders also enjoy

high abnormal returns in tender offers, reporting a positive 1.2 percent abnormal returns in

tender offers, higher than those in other offers (-3.6%).

Many additional studies have provided explanations of the fact that abnormal returns in tender

offers are higher than those of mergers. Bradley et al. (1988) suggest positive target returns in

tender offers are a result of increased competitiveness of the market, reasoning that the

introduction of the William Act in 1968 is the main driving force. However, this view has been

challenged by Franks and Harris (1989), who found positive returns to the U.K. targets both

before and after 1968. Martin (1996) and Dong et al. (2006) suggest that tender offers are likely

to be those acquisitions financed by cash that conveys the market good news. Kohers et al.

(2007) suggest bidders’ propensity for using tender offers is larger in the cases where targets

are defensive and there are multiple bidders. This line of studies suggests that bidders can better

explore synergies when proposing tender offers.

2.3.3.6. Size and relative size

The size of each firm involved in an M&A deal is also an important factor influencing M&A

performance of the firm. Agrawal et al. (1992) were the first to account for the size of the firm

when examining the bidder long-term abnormal returns, as there are a large number of large

firms in their sample. Further to this, Moeller et al. (2004) find bidder size is a key factor in

explaining bidder’s performance. These authors examined the relationship between bidder size

and firm performance in the M&As and find that managers of large bidders tend to destroy

more of the firm’s value than their smaller counterparts, reasoning that managers of large firms

who do not have any restrictions regarding using the firm’s resources potentially dissipate these


73

resources in M&As. Evidence that bidder size and bidder gains are negatively correlated is also

supported in prior M&A studies (Eckbo and Thorburn, 2000; Moeller et al., 2005).

In addition, target size is also a crucial factor in assessing M&A performance. Analysing a

sample of 3,691 U.S. mergers between 1990 and 2007, Alexandridis et al. (2013) find a

negative relationship between target size and bidder gains. In particular, announcement returns

to large targets were 2.37 percent lower than small targets and bidder announcement returns

decrease by 1.1 percent for every standard deviation increases in target size. The authors

explain this as post-merger integration problem of large deals, as they are complex, hence, they

are hard to create value for the firm. According this theme in the literature, relative size as well

as the size of the two firms are essential in examining M&A performance of the firm and should

therefore be included (Agrawal et al., 1992; Loderer and Martin, 1992).26

Relative size is the deal value scaled by the bidder value, which captures the materiality of the

deal decision to the bidder. A larger ratio indicates that the M&A decision is relatively more

important to the bidder. Relative size measures the value of the two firms involved in an M&A

deal, exhibiting a substantial impact on M&A outcomes. Asquith et al. (1983), by examining

the effect of relative size of the target and bidder on bidder gains, noted a very small target firm

relative to the bidder does not increase the value for the bidder. While a target firm whose size

is about half of the bidder generates 1.8% higher abnormal returns for the bidder than the target

whose size is only one-tenth of the bidder. In line with this, Jarrell and Poulsen (1989) argue

that a small target relative to the bidder may fail to raise the market’s attention, leading to

insignificant market reactions to the bid announcement, which is supported by Loderer and

Martin (1990) who include private acquisitions in their M&A sample.

26 These authors adjust for size of the firm in assessing a firm’s long-term stock returns.


74

2.3.3.7. Market-to-book ratio

Market-to-book ratio (MTBV) reflects the firm’s prospects, which is also a crucial factor in

assessing M&A performance of the firm. In a core paper, Rau and Vermaelen (1998) find that

the post-merger performance of the glamour bidder (i.e. firms with higher MTBV) is

significantly worse than that of the value bidder (i.e. firms with lower MTBV) regardless of

the choices of mergers and payment methods. The authors attempt to justify this finding with

three hypotheses: the performance extrapolation hypothesis, the method of payment hypothesis

and the Earnings Per Share (EPS) myopia hypothesis.

Analyzing a comprehensive U.S. acquisition comprising 3,169 mergers and 348 tender offers

between 1980 and 1991, Rau and Vermaelen (1998) found that glamour bidders underperform

value bidders in a three-year period following acquisitions. Their study showed that bidders in

mergers underperformed their match firms significantly by 4 percent while bidders in tender

offers enjoyed positive abnormal returns of 9 percent. These results suggest that the market

tends to over-extrapolate the past performance of the glamour firms. The market believes that

firms with high MTBV tend to have high past stock returns and high growth in cash flows and

earnings, leading to an overestimation of the firms’ performance around the announcement

date. It is suggested that the market belief to the firm’s performance enhances managerial

confidence and result in an overpayment, which leads to the same prediction of the hubris

management hypothesis. In addition, the authors revealed that glamour bidders prefer stocks

to cash as a means of payment for acquisitions, as stocks are overvalued by the market in the

presence of the market’s over-extrapolation bias. In the long term, market correction to

overvalued stocks downgrades the price, leading to negative abnormal returns. However, their

results do not provide any additional support for the prediction that the EPS is an important

determinant for the glamour bidders’ underperformance.


75

Sudarsanam and Mahate (2003) provide similar findings when examining a sample of

successful U.K. acquisitions between 1983 and 1995. They used two proxies for glamour/value

stocks: price-to-earnings ratio (PER) and MTBV and assessed the firm’s performance with

various benchmark models. Their results showed that the value bidders outperformed the

glamour bidders over three years in the post-merger period, arguing that underperformance in

the long term is due to the fact that the U.K. market forbids the market to over-extrapolate the

firm’s past performance.

Pachare (2010) explains the extrapolation phenomenon with the level of market power that a

firm possesses. He proposes that higher market power possessed by the firm relative to its

industry peers is associated with a stronger positive investors’ belief in the firm’s prospects.

Their study indicates that firms with high market power tends to increase the market’s over-

extrapolation, leading a stronger short-term momentum followed by stronger long-term

reversals.

On the whole, the studies cited above show that MTBV should be included in assessing the

M&A performance of the firm for both the short and long terms. Fama and French (1993),

criticizing the work of Agrawal et al. (1992), who do not adjust for MTBV while assessing the

firm’s M&A performance suggest that MTBV is able to capture a large proportion of the

average stock returns for the firm. Therefore, it is essential to include MTBV in an M&A study.

Chapter 2. Literature Review: Prospect theory

76

2.4. Prospect theory

Kahneman and Tversky (1979) were the first to propose prospect theory, which is a descriptive

theory of decision-making under risks. Such a decision-choice model decipts an S-shaped value

function, showing concave for gains and convex for losses. There is a diminishing sensitivity

in the loss and gain domains, and such a trend in loss domain is much steeper than in gain loss.

This leads to an implication that given an equal change in these two domains people tend to be

more sensitive to losses rather than gains, i.e. loss-aversion tendency. Since people care about

the loss more than the gain, they tend to be risk-seeking when experiencing losses and risk-

averse when enjoying gains. However, gauging gain and loss is subjective. The expected

expected utility27 theory (i.e. EUT) judges people’s feelings about the gain and loss with a zero

value, the positive outcome is described as a gain, whereas the negative outcome is regarded

as a loss. And the theory uses weighted average of outcomes to measure the expected value of

the choice.28

One of the leading features of prospect theory is that the gain and loss is determined by a

specific value point, termed as the reference-dependence bias. Even though the outcome is

postive, it can be regarded as a mental loss if it is below the expectation. In addition, unlike

EUT, which potrays the outcome of a decision, prospect theory puts much more emphasis on

a decision-making process, since it outlines how the change of the value affects the outcome

of a decision. Prospect theory has been widely applied to address various fields of study over

the decades. Barberis (2013) summarised three important implications of prospect theory:

diminishing sensitivity, loss-aversion tendency and reference dependence, and these were

derived from two types of evidence: human-related experiments and empirical evidence.

27 The utility is the value placed on outcome. 28 According to the EUT, The expected value of the choice is sum of the product of the probability and outcome

values of each probability.


77

A number of studies document empirical evidence in support of prospect theory. Northcraft

and Neale (1987) conducted an experiment on the estimates of house price. They found that

the experienced house agents estimate house prices in relation to a random house-listing price.

Loughran and Ritter (2002), using the discontinuity analysis, found that current CEOs raise

new equity in accordance with the share price at the time they started to charge the firm. Shefrin

and Statman (1985) observed that investors tend to sell winner stocks too early and keep the

loser stocks too long. In a similar vein, Odean (1998), by studying the detailed trading records

of 10,000 accounts from 1987 to 1993, which were collected from a large discount of brokerage

house, found that individual investors were less likely to sell loser stocks to realise paper loss,

indicating that the loss-aversion is a prevalent feature in the stock market.

2.4.1. Diminishing sensitivity

Kahneman and Tversky (1979) suggest diminishing sensitivities in both the loss and gain

domains. According to their experiment, subjects tend to be more sensitive when the

temperature changes from 3 degrees to 6 degrees than if it changes from 13 degrees to 16

degrees. This assumption also applies to the monetary situation where a loss of $10 has more

impact on a person with $100 than on a person with $1000. It is suggested that the change of

the value affects people’s feelings rather than the final states. Finally, the experiment also

documented that people have a loss-aversion tendency insofar as they are reluctant to realise

losses relative to the reference point. It has been explained that experiencing pain makes a

greater impression than experiencing pleasure.

2.4.2. Reference dependence

A key distinguishing feature between prospect theory and expected utility theory is the

reference-dependence. Prospect theory describes gains and losses on a bias of a given specific


78

point, which serves as the benchmark. Kahneman (2003) maintained that prospect theory

captures the decision-making process whilst the expected utility theory simply reflects an

outcome of the decision-making process, failing to address the effect of the change of the value

on people’s feelings. The reference-dependence preference also predicts that decision-makers’

personal judgements are subjective to an established level of reference, and they are satisfied

when returns are higher than the reference point and dissatisfied when the returns are lower.

Tversky and Kahneman (1974) conducted a series of human-subject experiments, affirming

that the estimates of participants are influenced by a random number in case of uncertainty.

Subjects were allocated in different groups and asked the percentage of African countries in

the United Nations. With insufficient information about this unfamiliar area, subjects randomly

estimated a number in relation to the arbitrary number provided by the experiment. The median

estimates for the percentage of African countries in the United Nations were 10 for the group

given the number of 25, and 65 for the group given the number of 45. Therefore, Tversky and

Kahneman identified the reference point effect on the decision-making process.

2.4.3. Loss-aversion tendency

Loss-aversion theory posits that investors generally have a tendency to not be willing to realise

loss (Kahneman and Tversky 1979), where the loss is relative to a reference point. In addition,

Tversky and Kahneman (1992) indicate a diminishing sensitivity suggesting the utility for

returns diminish in both gain and loss domains. The loss-aversion tendency has entered trading

behaviour as a disposition effect. Prior studies have put forward loss aversion theory in

different fields of finance. Using a realised gain and lose model, Barberis and Xiong (2009)

indicate the disposition effect to describe investors’ trading behaviour. Mehra and Prescott


79

(1985) and Benartzi and Thaler (1995) suggest that the disposition effect should provide an

explanation for the equity premium puzzle.29

2.4.4. Prospect theory and M&As

Baker et al. (2012) were the first to put forward the idea of applying reference point theory to

M&As. Studying a comprehensive sample of U.S. public acquisitions, the authors found that

the reference point price measured with the target 52-week high, has a substantial impact on

M&A participants’ decisions-making process. Their study indicates that the target 52-week

high serves as bargaining power for the target who negotiates with the bidder. Since the value

of the firm is hard to estimate and lacking time in which to process the information, target

shareholders tend to believe the target 52-week is their firm’s fair value and approve the deals

if offer price is paid according to this reference point price. Target managers who avoid being

accused of failure to meet the commitment to the firm attempt to justify the offer price with

this reference point price. This is similar to bidders who have to justify how much they should

pay for the targets to obtain the deal. According to Baker et al. (2012), the bidder may reason

that ‘if the target was valued at a certain level just a few months ago, shouldn’t we, with our

ability to realise synergies, value it near or above the same level?’, implying that the M&A

pricing decisions made is not only influenced by the reference point but also by the managerial

overconfidence, since bidders do not have any solid evidence regarding the target’s value.

Baker et al’s study (2012) showed that the target 52-week high is significantly positively

related to the offer premium and bidders receive negative market reactions around the takeover

announcement date when paying for acquisitions based on the target 52-week high. The authors

29 Benartzi and Thaler (1995), building on the study of Mehra and Prescott (1985), proposed the myopia loss-

aversion as a solution to the equity premium puzzle.


80

reason that targets are less likely to give up firms unless the offer premium can compensate

their mental loss, measured relative to a reference point price. The bidder whose aim is to

successfully obtain the deal is likely to pay a price relative to the target 52-week high to avoid

the loss-aversion of the target. According to the analysis of Baker et al. (2012) regarding the

relationship between deal success and the target 52-week high, it was found that paying

according to the target reference point increases the likelihood of deal succeeding. While, on

the other hand, bidder shareholders find a price relative to the target 52-week high as an

overpayment, since high offer premiums transfer their wealth to that of the target shareholders

and lead the managers hard to recover following the transaction, thus negatively react to the

bid announcement. Their findings are consistent with implications of prospect theory.

Following Baker et al’s idea (2012), Chira and Madura (2015) focused on takeover probability

and the reference point effect. Chira and Madura (2015) suggest that bidders should also assess

their value according to the bidder reference points, finding that optimistic managers who are

more likely to acquire their firms (i.e. a management buyout) than the outsiders when the firm’s

price is low relative to the 52-week high. In addition, their study documented that a firm with

the nearness 52-week high is probably the bidder, whereas a firm with the farness 52-week

high is probably the target in an M&A deal, indicating the role of the firm’s 52-week high

played as the firm’s bargaining power.

81

CHAPTER 3: REFERENCE POINT

THEORY ON CROSS-BORDER AND

DOMESTIC M&A DEALS: U.K.

EVIDENCE

Chapter 3. Reference Point Theory on Cross-Border and Domestic M&A Deals: U.K. Evidence

82

Abstract

There is no consensus on how much a target is worth except there are perceptions of the manager and

the market to the firm’s valuation. This chapter examines the reference point theory of M&As, a

behavioural finance theory that aims to explain what determines the decision of M&A pricing. The

researcher put forward Baker et al.’s idea (2012) of finding a strong reference point effect on various

key aspects of U.S. M&As in the U.K. market, where the reference point effect is expected to have

stronger explanatory power for offer premiums and the market reactions due to the unique features

of this market. In this chapter, the reference point effect was used to analyse offer premiums and

bidder announcement returns of a sample of 155 cross-border and 451 domestic acquisitions into the

United Kingdom. Using the target firm’s 52-week high as the reference price, an overall significantly

positive relationship between the reference price and offer premiums was established. In addition, a

reference point distinguishes the takeover motives of domestic bidders from cross-border bidders. It

was established that offer premiums driven by the target 52-week high lead to negative market

reactions for domestic bidders, suggesting evidence of overpayment. However, there is little evidence

of overpayment in cross-border bidders who pay in accordance with the target 52-week high. The

findings of this chapter are consistent with prospect theory.


83

3.1. Introduction

“Necessity never made a good bargain.”

Benjamin Franklin

Valuating a target firm is a difficult task for bidders. A sizable stream of M&A research has explained

the price that a bidder offers for the target from various perspectives of the bidder firm, such as the

synergy hypothesis and the overpayment hypothesis (Moeller et al., 2004; Alexandridis et al., 2012).

Under the synergy hypothesis, offer premiums can be viewed as an exchange of the expected

synergies following acquisitions (Antoniou, et al., 2008), while under the overpayment hypothesis,

offer premiums are a result of overpayments, which are largely because of managerial overconfidence

(Hayward and Hambrick, 1997; Malmendier and Tate, 2005).

The price that the bidder offers to the target is an outcome of the negotiation between the boards of

the target and the bidder, and such a price involves the perception of the managerial view of the firm’s

valuation. Since how large the M&A offer premiums are paid for the target determines the wealth

distribution between the target and the bidder shareholders, an offer cannot be made without an

agreement between the two firms involved. With this in mind, analysing takeover motives from either

the bidder or the target side alone does not fully account for the rationale behind the offer price.

In this chapter, the reference point theory of M&As was established to explain the offer price. The

theory was first developed by Kahneman and Tversky (1979), depicting how people make decisions

under risks. Putting this in the context of M&As, the theory sheds light on how managers make major


84

investment decisions facing complex situations. Apart from those behaviour finance theories that

explain M&A motives from the outcome of the deal, prospect theory allows use to directly observe

the managerial decision-making process of the offer premium.

Edmister and Walkling (1985) suggest that the offer premium reflects the bargaining power of the

management teams of the target and the bidder, implying that valuing a firm requires a series of

cognitive biases of the management team of the bidder and the target. Due to the complexity of the

M&A activity, the offer premium cannot be set precisely. Therefore, addressing the M&A valuation

effect requires further investigation from the perspective of managerial cognitive biases.

The reference point effect is an important implication of prospect theory developed by Kahneman

and Tversky (1979), affirming that people have a pre-determined piece of information in mind and

gauge losses and gains according to it, and people’s feelings about the loss and gain tend to diminish

according to the reference point. They are more sensitive to the value that is nearer the reference point

than far from it (i.e. diminishing sensitivities). In addition, prospect theory posits that people are

generally loss-aversion, in that they are not willing to realise loss relative to the reference point they

have settled upon. The implications of the reference point theory have been highlighted in

psychological research (Tversky and Kahneman, 1974) and also applied in financial and economic

research (Shefrin and Statman, 1985; Odean, 1998; Barberis and Xiong, 2009).

A firm’s 52-week high is publicly available information and serves as a salient reference price in

previous literature (Heath et al., 1999; Huddart, et al., 2009; Burghof and Prothmann, 2011; Baker et

al., 2012). In Tversky and Kahneman (1974), people tend to rely heavily on a single piece of


85

information while making decisions. The 52-week high has been widely reported by some financial

media, which shapes investors’ impression. In addition, investors are likely to be influenced while

lacking information. Burghof and Prothmann (2011) suggest that the 52-week high effect is more

pronounced when there is high level of information uncertainty.

Baker et al. (2012) proposed the reference point theory for M&As, examining how target valuation

could alter the offer premium. These authors were the first to put forward the idea of applying the

reference point effect on an M&A testing ground. The theory explains the offer premium from a

psychological perspective from both the bidder and the target sides. The authors indicate that the

reference point measured by the target 52-week high is positively related to the offer premium. They

find that bidders tend to offer a price for their targets relative to the target 52-week high, which

suggests that both targets and bidders are influenced by the reference dependence bias (i.e. the

reference point effect). In addition, it was found that the market tends to react negatively to the bidder

announcement when the offer price is paid according to the target 52-week high, which implies that

the target 52-week high is employed to reinforce the bargaining position of the target during the

course of the deal negotiation. If a high offer price destroys the wealth of bidder shareholders, why

bidders pay in accordance with the target 52-week high. On the other hand, what facilities the target

in using the reference point requires further examination.

This chapter provides further evidence in relation to the reference point effect on the offer premium

the U.K. market. The reference point effect in a comprehensive public acquisition sample of both

U.K. domestic bidders and cross-border bidders into U.K. public targets will be investigated. A

comprehensive public acquisition sample of both domestic and cross-border will enable assessment


86

of the role of the reference point effect plays in a different setting outside U.S. market. Previous

studies show the level of bidding competition is more intensive in the U.K. takeover market than any

other takeover market in the world (Moeller and Schlingemann, 2005). Alexandridis et al. (2010)

found that higher level of competition is accompanied by positive target announcement returns,

higher offer premiums and negative bidder announcement returns. Their findings suggest that public

targets in a more competitive market have relatively greater bargaining power in M&A activity, and

are likely to translate it into higher offer premiums based on their 52-week high. Targets in such an

environment are more likely to use a reference point price to negotiate for a possible high price from

the bidder. Therefore, the researcher hypothesises that the reference point theory for M&As is

pronounced in the U.K. market, which is also competitive.

The researcher contributes to this thread of literature by directly examining the role of reference point

theory in the setting of cross-border M&As of U.K. targets. Erel et al. (2012) and Uysal et al. (2008)

affirm that the likelihood of acquisitions increases when the target nation is closer to the bidder nation,

which suggests geographical distance impedes information transmission in matters of investments.

Domestic bidders are supposed to have informational advantages over cross-border bidders due to

the sophisticated local networks. Therefore, it is expected that cross-border bidders lacking

information about their targets are more likely to put more weight behind publicly accessible

information such as the target 52-week high. Therefore, bidders might have to offer a premium based

on the target 52-week high.30

30 Croci et al. (2012) provided an alternative explanation regarding information asymmetry and offer premiums, arguing

that bidders with favourable asymmetry information tend to offer higher premiums. Bidders can therefore be expected to

regard the target 52-week high as its potential profit-generating ability, and offer high premiums when the target 52-week

high supports the favourable asymmetry information they might have.


87

Furthermore, strong investor protection in the U.K. market could potentially reinforce the reference

point effect. Under the shareholder-oriented legislation environment, U.K. target managers are

unlikely to pursue private benefits at shareholders’ expense (Armour and Skeel, 2007). Wulf (2004),

however, found U.S. evidence that agency conflicts between the target manager and the shareholder

result in low offer premiums as target managers accept cheap offers associated with larger personal

benefits. Therefore, the reference point effect should be more pronounced in markets with stronger

investor protection than those without. In this pursuit, the U.K. market also provides an ideal testing

ground for the reference point theory.

Accounting for the majority of the deals in the European countries, the U.K. takeover market is the

second largest market in the world (Faccio and Masulis, 2005; Croci et al., 2010; Gugler et al.,

2012).31 Recent decades have seen a rapid growth of cross-border bidders’ acquisitions of U.K. public

targets. Many additional studies focus on the cross-border M&As into the U.K. market. Danbolt (1995)

attributes the U.K. takeover boom in the late 1980s to the growth of cross-border acquisitions.

Following his study, Gregory and O'Donohoe (2014) analysed 119 cross-border bidders into the

United Kingdom over the period 1990-2005, where the cross-border acquisitions took up to 41% of

their U.K. public acquisition sample.

31 By studying comprehensive European acquisitions announced between 1997 and 2000, Faccio and Masulis (2005)

reported a large proportion of deals involving U.K. bidders (65.3%) and U.K. targets (47.0%). Croci et al. (2010) indicate

that U.K. takeover market is an active takeover market which accounts for 65% of mergers in Europe. By studying the

U.K. completed mergers during the fifth merger wave, Gugler et al. (2012) found that the U.K. experiences a similar

takeover trend compared with the United States. In the U.K. M&A market, the overall transaction value reached $360

million in 2000. At the same time, the total transaction value for the M&A market in other major European countries was

$260 million.


88

Given the strong investor protection in the U.K. market, cross-border M&As directed at U.K. targets

provide a natural hedge against any economic and political instability for the bidder. To the best

knowledge of this researcher, this is the first study using a cross-border M&A sample to test the

reference point theory of M&As in the U.K. market. The findings of this chapter imply that stronger

investor protection enhances the target bargaining power, increasing the probability of the target

requires higher offer premiums from the bidder based on the target 52-week high.

The results of this chapter are consistent with the implications of prospect theory. This chapter yields

two main findings. First, it became apparent that the target 52-week high is significantly positive

relative to the offer premiums, suggesting that an offer price linked to the target 52-week high is

perceived as a reference point for the target shareholders in both domestic and cross-border M&As.

With the prospect theory framework, targets are generally loss-aversion, in that they are reluctant to

accept an offer that is significantly below the reference point, hence pushing their managers to

negotiate for higher offer premiums. Targets find it easier to require high offer premiums from the

cross-border bidders than from the domestic bidders. However, it was found that bidders no longer

pay based on the target 52-week high when the current price deviates greatly from it. It is possible

that the target’s current price is less likely to reflect its best performance in the recent past, leading

bidders to walk away. Our results therefore imply that bidders are bounded by rationality.32

Secondly, the market reacts negatively to the bid announcement in the sample consisting of both

domestic and cross-border M&As, which suggests that the market views an offer price based on the

32 Despite the fact that bidders make decisions based on the superficial information they possess they have rationality of

preventing things from going wrong.


89

target 52-week high as an overpayment. While investigating the role of the reference point effect in

these two subsamples and using the target 52-week high as an instrumental variable of the offer

premium, it was established that the offer premium driven by the reference point effect does not make

any greater difference to the bidder announcement returns under the 2SLS estimator than by OLS in

the cross-border M&A sample. It does, however, lead to more negative bidder announcement returns

under 2SLS in the domestic M&A sample. These results show in the case of domestic bidders who

offer a price according to the target 52-week high tends to be perceived as overpayment, whereas the

market has already taken the reference point effect into consideration for bidders making cross-border

M&As.

Two distinct contributions have been made to M&A literature. First, to the best of the researcher’s

knowledge, this is the first study to test the reference point effect using cross-border M&A sample.

The cross-border M&A sample was used as a control sample for the domestic M&A sample to

investigate the reference point effect in a setting where there exists larger information disadvantages.

It is expected that the reference point effect more pronounced among cross-border bidders given they

have a less sophisticated information networking than their domestic counterparts. Therefore, the

reference point theory helps explain how bidders value targets in an environment of information

uncertainty.

Second, the reference point theory distinguishes the motive of domestic bidders from cross-border

bidders acquiring U.K. public targets. The offer premium based on the reference point has a

significantly negative impact on domestic bidders’ announcement returns yet an insignificant impact

on cross-border bidders’ announcement returns. It can be argued that shareholders of domestic


90

bidders may believe that managers pay too much for the target, based on the target 52-week high, and

thus react negatively to the bidder announcement, whilst those of cross-border bidders may take the

target 52-week high for granted.

The remainder of this chapter is organised as follows: Section 2 provides a literature review related

to the offer premium, prospect theory, and cross-border acquisitions. Section 3 develops hypotheses

with regard to the reference point effect. Section 4 summarises our data and presents the methodology

used in this chapter. Section 5 analyses our results. Section 6 concludes this chapter.

3.2. Literature review

3.2.1. Cross-border M&As in the U.K.

The level of cross-border acquisitions in the U.K. has grown rapidly over the past three decades.

Danbolt (1995) held that the U.K. takeover boom during the 1990s was largely because of the increase

of cross-border acquisitions. His study documents that cross-border acquisitions into the U.K.

accounted for 58% of the total value of this market in 1990. Erel et al. (2012) conducted a study into

a sample of 56,978 cross-border acquisitions in 48 countries between 1990 and 2007, taking up one

third of worldwide acquisitions. It was found that the volume of cross-border acquisitions increased

from 2002 to 2007, reaching to 45% of all M&A activities up to 2007. Of particular note, U.K. targets

were engaged in 6,753 cross-border acquisitions.

The level of U.K. M&A activity has turned the market into an active takeover market. Faccio and

Masulis (2005) in a sample of European M&As over the period of 1997 to 2000 reported that a 47%

of acquisitions involved U.K. targets. Consistent with their study, Croci et al. (2010) found that the


91

level of U.K. M&As took up 65% of takeover activities in Europe between 1990 and 2005, suggesting

that U.K. is the most active takeover market in Europe. Moeller and Schlingemann (2005) argue that

the U.K. takeover market was even more active than the U.S. for its strong shareholder right

protection and strict legislative environment. Alexandridis et al. (2010) ascribe the activity of the U.K.

takeover market to intense competition. Their study shows that the number of competing bids in the

U.K. was more than that in the United States.

Research has identified the motives of cross-border acquisitions. For example, Danbolt (2004) noted

that cross-border acquisitions may not be simply driven by value maximisation as to the same degree

as domestic acquisitions. Instead, bidder managers are prone to increase their power, status and salary

through cross-border acquisitions. Doukas (1995) and Doukas and Travlos (1988) argue that cross-

border acquisitions play a value-enhancing role as they integrate the imperfect capital markets across

countries, while Erel et al. (2012) noted that cross-border acquisitions were driven by the high quality

of governance standards. Their worldwide evidence shows that in countries with a higher quality of

corporate governance, accounting standards have stronger shareholder protection for the combined

entities. Rossi and Volpin (2004) provided some similar findings, showing that the volume of M&A

activity is significantly larger in countries with better accounting standards and stronger shareholder

protection. Their findings imply that cross-border acquisitions act as a tool for a country with a

substandard legislative environment to access a country with a better legal environment.

One particular U.K. study by Danbolt (2004) identified the international risk diversification

hypothesis as the most plausible explanation regarding the motive of cross-border acquisitions in the

the U.K. compared with the market access hypothesis and the exchange rate hypothesis. Based on a


92

sample of 514 domestic and 116 cross-border acquisitions in the U.K. setting over the period between

1986 and 1991, Danbolt found that targets gained significantly higher abnormal returns in the cross-

border acquisitions than in the domestic acquisitions. In addition, his study shows that geographical

distance between the bidder country and the target country affects the wealth effect of the target

shareholder. The target abnormal returns were found to be significantly higher when bidders are based

outside the U.K. market. However, his results were not robust when different event windows were

employed in his study.

Despite the fact that substantial research has identified the benefits brought by cross-border

acquisitions, valuing a target is a more difficult task for cross-border bidder managers than for

domestic bidder managers given different accounting standards, cultures and a large amount of

asymmetric information country by country. Danbolt (2004) noted that cross-border bidders were

prone to pay higher offer premiums for their targets due to the size of valuation error caused by these

differences between countries.

Uysal et al. (2008) in their study of the relationship between geography and bidder returns in the U.S.

context found that bidders tended to gain higher returns when acquiring a closer rather than a farther

target, which is because of the information advantage obtained from geographical proximity. The

authors argue that information is likely to be coded, transmitted, and interpreted in a nearer country

rather than a more distant one. In addition, local managers are likely to obtain information through

their social networks.


93

By studying an international cross-border acquisition sample, Erel et al. (2012) found the likelihood

of acquisitions increases when a bidder country is closer to its target country. Their results show

significantly positive coefficients of geographical proximity at a 1% level to explain the bidder returns.

Together, bidders engaging in cross-border acquisitions may encounter information barriers that

potentially cost the bidder and benefit the target.

3.2.2. The reference point effect on the U.K.

While most work has focused on the U.S. market, it is interesting to explore the reference point effect

on cross-border acquisitions in the U.K. public targets. First, the high level of competitiveness in the

U.K. market entitles its targets to greater bargaining power, which makes the reference point effect

pronounced. Danbolt (2004) affirm that the cross-border target effect means that U.K. targets are

outperformed in cross-border acquisitions rather than in domestic counterparts due to risk

diversification. In line with his work, Alexandridis et al. (2010), by investigating a sample of global

M&A activity over the period of 1990 and 2007, found that the level of competition increases with

target abnormal returns and the offer premium, and decreases with bidder abnormal returns. Their

results show that when the market is more competitive, targets are more likely to translate their

bargaining power into benefits. It is generally believed that cross-border bidders have less information

advantage than domestic bidders, which reduces their bargaining power over their targets. With this

in mind, U.K. targets are more likely to ask for an offer price based on a reference price favouring

their shareholders from cross-border bidders than domestic bidders.

Secondly, less severe of agency conflicts between the manager and the shareholder in the U.K. could

potentially enhance the reference point effect. Although it is generally believed that the corporate


94

governance system in the U.K. is consistent with the U.S. (La Porta et al. 2000), yet the U.K. has a

better corporate governance system in relation to M&As compared with the US. Armour and Skeel

(2007) in their review of U.S. and U.K. takeover regulations conclude that the U.K. has self-regulation

takeover regulations which are strongly orientated towards the shareholder’ interests, whereas the

U.S. law system entitles managers to absolute discretionary power over the firms. In addition, in the

context of takeover defences the authors suggest that the U.K. takeover code implicitly bans any

“frustrating actions” against their shareholders, which is different from the United States. Thus,

takeover regulations in the U.K. can potentially reduce the conflicts of interests between the manager

and the shareholder. The bargaining power of target managers is enhanced under more stringent

regulations and target managers can require higher offer premiums from outward bidder managers in

M&As.

The corporate governance system is firmly rooted in the legal environment of a country. Research

has established some contradictory findings in relation to the divergence of the legal environment

effect on shareholder valuation in cross-border acquisitions. Moeller and Schlingemann (2005)

analysed bidders’ acquisitions of the worldwide targets over the period of 1985 and 1995 and found

that bidders tend to pay higher offer premiums to targets with weaker legal environment compared

with the United States. Thus, targets with more information asymmetry and poorer corporate

governance system tend to accept lower offer premiums, which leaves the bidders opportunities to

explore the benefits.33

33 Addressing the U.K. market, Antoniou et al. (2008) noted that the lower offer premium does not benefit bidders.

However, their study does not include cross-border acquisitions.


95

Hagendorff et al. (2008), analysing the bidder returns of financial firms between 1996 and 2004 in

the context of the U.S. and Europe, found that target countries with a weaker legal environment are

more likely to accept fewer offer premiums. Unlike Moeller and Schlingemann (2005), Hagendorff

et al. (2008) argued that information asymmetry and severe agency problems lead to a decrease in

the offer premium.

Martynova and Renneboog (2008) studied a sample of listed bidders’ acquisitions of both listed and

unlisted targets in Europe during the fifth M&A wave. They revealed bidders obtain positive returns

when acquiring targets regardless of their legal environment. They ascribed targets in stronger legal

environment compared with bidders to the bootstrapping effect that bidders have incentives to get

access to better corporate governance system. Based on their study, targets in a strict legal

environment are expected to have greater bargaining power than those in a less strict legal

environment.

On the whole, studies have highlighted how the legal environment impacts the bargaining power of

both the bidder and the target in the context of M&As. However, results are mostly driven in the U.S.

or the European sample in different sample periods (Moeller and Schlingemann, 2005; Hagendorff et

al., 2008, Martynova and Renneboog, 2008). Little has been done relating to the issue of the U.K.

market. Given that the U.K. is a more active and competitive market than that of any other country in

the world due to the stringent shareholder protection and legal environment (Moeller and

Schlingemann, 2005; Alexandridis et al., 2010), target managers are entitled to greater bargaining

power and tend to demand higher offer premiums accordingly. Bidders from countries with relatively

a weak legal environment compared with the U.K. are more likely to pay higher offer premiums in


96

exchange for stronger shareholder protection. With a higher level of information asymmetry, bidders

are more likely to be misled by the value of the target, thus valuing targets based on readily available

information. Results regarding the reference point effect being more pronounced on cross-border

M&As.

3.2.3. The target 52-week high as a reference point price

Over the years, attention has focused on a firm’s 52-week high as a reference price. The 52-week

high is widely reported by a range of media on a daily basis, such as the Wall Street Journal, Yahoo

Finance, Financial Times, or the South China Morning Post, leading investors to gain full access to

information and may use it to assess the firm’s performance.

Prior research has identified the 52-week high’s salient role as the reference point in the decision-

making processes. Baker et al. (2012) applied the reference point theory in a comprehensive sample

of U.S. public acquisitions, documenting that the reference price measured with the target 52-week

high had a substantial impact on M&A decisions-making process. The target 52-week high is found

to be significantly positively related to the offer premium and bidders receive a negative market

reaction around the takeover announcement date when paying for acquisitions based on the target 52-

week high. The authors explained it with the reference point theory, reasoning that targets are less

likely to give up firms unless the offer premium can compensate target shareholders’ mental loss

relative to the reference price. Their statement implies the reason that bidders pay in accordance with

the target 52-week high is to avoid target shareholders’ loss-aversion, and thus increase the likelihood

of deal completion. It was found that shareholders negatively react to the bid announcement since


97

they believe paying based on the target 52-week high is too high. Baker et al’s findings (2012) are

consistent with the implications of prospect theory.

The 52-week high is significant in explaining investors’ behaviour in a variety of studies. Heath et al.

(1999), for example, found a relationship between individual stock option exercise decisions and the

reference point. Their study includes stock option exercise decisions made by over 50,000 employees

at seven publicly traded corporations and finds that the number of employee stock option exercise

activities nearly doubles when the stock price exceeds the maximum price of the previous year. The

authors reason that individual investors have a reference price relative to the maximum stock price

when exercising stock options.

Huddart et al. (2009) provided evidence in a large sample that there is high trading volume around

the past price extremes. Using the weekly observations of 2,000 firms spanning 24 years from 1982

to 2006, the authors found that the past price extremes are significantly related to the trading volume

at the 1% level. The results continue to hold after controlling for a series of variables that have

significant impact on the trading volume addressed in earlier studies. The authors attribute the

salience of the past price extremes to bounded rationality and attention hypotheses that individual

investors tend to rely heavily on past price extremes, since they are limited in obtaining and analysing

information. Their study complements evidence that the 52-week high has a strong effect on the

trading decisions of the investor.

In a similar vein, Burghof and Prothmann (2011) suggest that the 52-week high effect can be

explained by the anchoring bias hypothesis, affirming that the 52-week high has strong predictive


98

power when there is a high level of information uncertainty. Investors can obtain profits under a

greater level of information uncertainty by focusing on the 52-week high strategy. Their results

indicate significant reference point effect on the trading behaviour of the investor.

The use of a certain piece of information as a reference point has been well documented in Tversky

and Kahneman's work (1974). It has been well known that human minds are likely to focus on a single

event and overlook other events that lead to different consequences when making decisions. Their

concept has been put forwarded into various studies and the firm’s 52 week high has been widely

adopted as a reference point candidate for its simplicity. Following these studies, the investor’s

decisions in conjunction with the firm’s 52-week high were measured.

Following Tversky and Kahneman’s work (1974), Rau and Vermaelen (1998) suggest the over-

extrapolation hypothesis indicating that investors put considerable weight behind the past

performance of the firm and attribute a higher expectation to the firm’s M&A performance when it

achieves a decent past performance, whilst they may underreact to a firm with a worse past

performance. Their study suggests that investors have a tendency to make predictions by exploring

the firm’s past performance. Based on the assumption that the market is not efficient, investors’ biases

on the past price information are likely to come into play in estimating the firm’s future performance.

3.2.4. M&A offer premiums

M&A offer premiums are an important factor that have received great deal of attention from

shareholders, as it directly determines how the value of the deal is distributed between the bidder and

the target shareholders. Bidders who pay high offer premiums for the target firms tend to signal to


99

their short-horizon shareholders that the firm is hard to generate wealth in the short run even though

it is possible for the firm to create value in the long term. On the other hand, target shareholders

benefit from high offer premiums. With this in mind, a higher offer premium is generally associated

with negative market reactions to the bidder firms but with positive market reactions to the target

firms. Prior literature on the offer premium has largely focused on the two completing hypotheses:

the synergy and the overpayment hypotheses. The synergy hypothesis tends to rationalise the motive

of bidders paying for the target firm as being to create value for the firm, suggesting that managers

serve the best interests of their shareholders should finding positive NPV projects, and premiums paid

for a target are in exchange for potential earnings to be created by a combined entity. The

overpayment hypothesis suggests that bidders are overconfident or the agency problems of the firm

are severe, leading managers to offer the target higher premiums in exchange for the control of the

target firm.

To justify the motive of offer premiums, Berkovitch and Narayanan (1993) tested three hypotheses

of M&A motives in a U.S. study. Their work provided a way of distinguishing synergy from

overpayment by looking at the correlation between target and total gains. It is suggested that

correlation should be positive if the bid is solely driven by synergy but negative if the bid is solely

driven by agency problem, and zero if the bid is solely driven by hubris. Their results indicate that

synergy and agency problem primarily explain the motives of U.S. M&As, with which Hodgkinson

and Partington (2008), agreed in their study of a sample of U.K. M&As over both the short- and long-

term windows, finding that bids are largely motivated by synergy and hubris, and that there is weak

evidence of agency-motivated bids.


100

Re-addressing the offer premium in the U.K. market, Antoniou et al. (2008) found that the synergy

hypothesis has more explanatory power than the overpayment hypothesis. Their study is based on a

sample of 396 successful U.K. public acquisitions between 1985 and 2004. The authors divide the

bidder sample into the higher, medium, and lower offer premium sub-portfolios, and conduct both

the short- and long-term analysis for each sub-portfolio. It was found that the short-term combined

abnormal returns (i.e. synergy) were positively correlated with the offer premium, and the long-term

abnormal returns to bidders did not show any significant differences between the higher and the lower

offer premium sub-portfolios. It is therefore suggested that synergy is a more plausible motive than

overpayment in explaining the offer premium in the United Kingdom.

Another area of literature addresses the offer premium with the overpayment hypothesis. In earlier

research, Roll (1986) found that hubris-infected managers are engaged in value-destroying bids.

Reviewing a series of U.S. and U.K. studies, the author concludes that managers are accused of

overpaying targets. They tend to believe that they have better-than-average skills which lead them to

dominate the market. As a matter of fact, neither the bidder nor the target managers can make correct

predictions as to a firm’s performance around the takeover announcement date. Their results show

that the combined entity generates significantly negative abnormal returns following acquisitions,

reasoning that managers are irrational and the market is efficient, thus managers are unable to explore

mispricing of the market and thus overpaying for their targets due to an overestimation of the expected

synergies.

In the light of this, Hayward and Hambrick (1997) investigated a sample of 106 large acquisitions

and found evidence of hubris-infected CEOs. These authors focused on large acquisitions


101

representing important decisions for the firms, tending to expose their CEOs to extensive risks. Their

study shows that bidders who achieve a decent performance in the recent past receive praise from the

media, together with higher salaries from their firms, which become the most important factors

leading them to believe that they are important in the firm. Thus, they are affected by hubris while

making major investment decisions. Their study suggests that the bidder’s recent performance, the

recent media praise for the CEOs, the CEO’s self-importance, and the composite index of these three

factors are positively correlated with the size of the offer premium, incurring negative market reaction.

Hayward and Hambrick's findings (1997) suggest that higher offer premiums are paid by hubris-

infected CEOs.

Using a comprehensive sample of U.S. M&As over the period of 1980-2001, Moeller et al. (2004)

established a positive relationship between the offer premium and the value of the bidder firm. The

authors provided various plausible reasons to explain why larger bidders tend to overpay for their

targets than their smaller bidders. First, larger firms generally have less concerned ownership than

smaller firms, leading to a weaker link in the interests between the manager and the shareholder.

Secondly, managers in a large firm are likely to be more overconfident as they are able to allocate

resources more easily, without any obstacles. Thirdly, larger firms are more constrained in finding

growth opportunities than smaller firms, thus managers are more likely to dissipate resources in

unnecessary projects, which destroy the wealth of their shareholders. Their findings indicate that

bidder managers of a larger firm, due to greater autonomy, are likely to be more overconfident than

those of a smaller firm.


102

However, Alexandridis et al. (2013) found little evidence that value-destroying acquisitions are

caused by large M&A offer premiums to the target shareholders and reason that the negative source

of the firm’s value is the neglect of the M&A integration process. They find an overall negative

relationship between the target size and the offer premium. Their study reported that the mean

(median) premium for the larger target is 38% (32%), whereas the mean (median) premium for the

smaller target is 54% (45%). Though a lower offer premium is paid to the larger target, it does not

create value for the firm. The authors assert that larger targets are value destroying for bidders, due

to bidders’ neglect of the target post-merger integration.

Incorporating both the synergy and overpayment hypotheses into the M&A study and using the

quadric model for a sample of 49 banking industry M&As between 1995 and 2004, Díaz et al. (2009)

found both the synergy hypothesis and the overpayment hypothesis come into play in their M&A

sample. More specifically, the authors suggest a non-linear relationship between the offer premium

and bidder abnormal returns. The lower level of the offer premium paid to the target shareholders

tends to create synergies for the firm, whilst the value of the combined entity is destroyed when the

offer premium further increases. Based on these results, the authors suggest the synergy hypothesis

is able to explain the lower level of the offer premium and the overpayment hypothesis is able to

explain the higher level.

Many additional studies have largely explained the offer premium in relation to these two competing

hypotheses. Bowman and Richards (2013) suggest that a higher offer premium is used in exchange

for the market power of the bidder, which generates synergy following acquisitions. Soegiharto (2009)

in his study of 3184 mergers between 1990 and 2010 finds no evidence that CEO hubris affects the


103

offer premium. By studying 81 European banking M&A over the period of 1994 to 2000, Diaz and

Azofra (2009) found that bidders tend to pay more for targets that have attractive characteristics, such

as it is larger size and has a decent recent performance. These authors argue that managers can have

substantial private benefits when acquiring another bank successfully, indicating that agency

problems are a source of overpayment. By contrast, Kim et al. (2011) argued that the offer premium

is to reflect upon the bank managers’ willingness for future growth. In that case, there are no conflicts

of interests between the manager and the shareholder and the primary purpose of paying M&A offer

premiums is to exchange for the control of the target firm.

3.2.5. Reference point price and the offer premium

Prospect theory offers new insights into the M&A offer premiums. It addresses the fact that prices

offered to the target shareholders not only originate from market reaction but also provides many

additional considerations to explain the behaviour of the main M&A participant. It differs from those

behavioural finance explanations relating to deal synergies and provides access to the market looking

at the negotiation process of the two firms. Since the offer premium is the outcome of negotiation

between the boards of the two firms involved, it reflects how the two firms think about the deal during

the course of the negotiation. Given a lack of information to observe the negotiation process of the

two firms, a high offer premium may signal to the market that the takeover target tends to dominate

the negotiation table and it is able to demand a price favouring their shareholders. Bidders, on the

other hand, will defend their position by providing market evidence that the target has the potential

to create synergies following an M&A deal, thus rationalising the M&A offer premiums. How much

a bidder firm should offer for the target firm is far too complex, whilst a single salient piece of public


104

information is more likely to impress the investors’ minds. By borrowing this, the bidder tends to

rationalise his/her motive for paying for the target and avoids being blamed on the market.

Baker et al. (2012) found that bidders tend to offer the target a price that is based on the target 52-

week high, indicating that bidders believe that they can beat the target 52-week high and thus create

synergies. However, their results show that the offer premium driven by the target 52-week high is

negatively related to the bidder announcement returns. The authors studied the role of the target 52-

week high plays in M&As. By doing so, they examined the OLS regression results for the bidder

announcement returns on the offer premium and compare these with the 2SLS regression results for

the bidder announcement on the offer premium, where the target 52-week high is used as an

instrumental variable of the offer premium. Baker et al. expect less negative or even positive market

reactions to bidders under the 2SLS compared with those by the OLS if the target 52-week high

reflects higher synergies and a more negative or a less positive bidder announcement returns if the

target 52-week high indicates overpayment.

Baker et al. (2012) suggest the role of the target 52-week high is an overpayment. Specifically, the

coefficient of the offer premium predicted by the OLS estimator is -0.040, suggesting that an increase

of 10% for the offer premium leads to a 0.4% decrease in the bidder announcement returns, while

using the target 52-week high as an instrumental variable of the offer premium. So, when using the

2SLS estimator, the coefficient for the offer premium becomes -0.245, suggesting that the target 52-

week high leads to more negative bidder announcement returns. Their results indicate that the market

believes the bidders’ offer to the target, whose current price is far below its 52-week high, is an

overpayment whereas the bidder may believe such a target leaves the firm more room for synergy


105

creation. Therefore, the manager and the shareholder tend to interpret this certain piece of price

information differently.

3.3. Hypothesis development

According to prospect theory, people tend to gauge gains and losses relative to a reference point, and

are unwilling to realise mental loss relative to this particular point. Since valuing a target is a difficult

task in M&As, the 52-week high is likely to be used for its simplicity. Burghof and Prothmann (2011)

suggest the use of a reference price increases with the level of information uncertainty. Moreover, the

52-week high reflects a firm’s recent best performance which provides predictions for the future

performance (George and Hwang, 2004). Finally, based on some experimental evidence from

Tversky and Kahneman (1974) and Kahneman and Tversky's works (1979), investors tend to rely on

a piece of information while deciding an atmosphere of risk and uncertainty. In the meantime, the 52-

week high as a reference price has been drawing the attention of the investor through a massive report

by the financial media.

The use of the reference point can be explained by the psychological influence of both the bidder and

the target. Target shareholders may find it hard to gauge managerial performance in M&A activity,

since they do not know for how much exactly their firm should be sold. Lacking information and time

in which to process the information related to the deal, target shareholders are likely to estimate their

firm based on a straightforward relevant price measure, such as the 52-week high. With a strong loss-

aversion tendency, they are not willing to see their firm, sold at a price significantly below their 52-

week high.


106

The psychological influence among the target shareholders has a substantial impact on the decision-

making processes of their manager. If the target manager sells a firm at a price far below the 52-week

high without a rational motive, the shareholders must blame their managers by arguing that they

would have been better off if they had not sold the firm. Managers thus may incur legal risks. The

magnitude of the premium received by the target manager can be seen as a judgement of whether or

not the manager aims to protect the wealth of their shareholders. It has been shown that U.K. has a

high quality of corporate governance system, which leads to less severe conflicts of interests between

the managers and the shareholders. In this stringent legal environment and shareholder protection,

the wealth of the target shareholders is the first priority. Target managers would bargain for the

premium based on the 52-week high, which is a visible price for the market. Baker et al. (2012)

indicate that bidders tend to pay a price based on the target 52-week high if it is closer to the target

current price, arguing that it might be a true reflection of the real performance of the target. Together,

a positive correlation between offer premiums and the reference price measured by the target 52-

week high in the U.K. market is expected.

H1a: There is a positive correlation between offer premiums and the reference price in the U.K.

market.

According to Erel et al. (2012) and Uysal et al. (2008), cross-border bidders have greater

disadvantages in gathering information than domestic bidders due to geographical distance. As the

level of information asymmetry increases, cross-border bidders are less likely to accurately estimate

the true value of the target, whereas domestic bidders tend to collect exclusive information from their


107

local social networks. In such a case, cross-border bidders are more likely to measure firm valuation

based on the target 52-week high.

In addition, a more competitive market environment tends to increase target bargaining power, thus

enhancing the reference point effect. Targets may find it easier to negotiate the offer premium based

on their 52-week high with the cross-border bidders than with the domestic bidders. In Edmister and

Walkling (1985), the offer premium is the outcome of negotiation. In the light of this, targets could

require higher premiums when they have greater bargaining power. Therefore, a stronger reference

point effect on cross-border acquisitions toward U.K. targets is expected.

H1b: There is a positive correlation between offer premiums and the reference price when cross-

border bidders acquire U.K. targets.

Moeller et al. (2004) suggest that overpayment increases the wealth of the target shareholder and

decreases the wealth of the bidder shareholder. The target 52-week high is perceived as the recent

best performance of the firm. Bidder managers are expected to possibly a lower price so as to increase

the wealth of their shareholders. Bidder shareholders may simply believe their managers pay too

much if their payment is based on the target recent best performance. It is generally believed that an

overpayment leads to a negative market reaction since bidder shareholders may believe such

premiums paid to the target are hard to recover following acquisitions. There, bidders’ performance

around the announcement date are expected to be negatively correlated to the offer premium when

deals are made with regard to the reference price.


108

H2: Bidder short-run performance is negatively correlated to offer premiums when deals are made

with regard to the reference price.

3.4. Data and methodology

3.4.1. Data

The initial sample contained 4,324 acquisitions of U.K. public firms announced between January 1,

1985, and December 31, 2014, extracted from Thomson One Mergers and Acquisitions Database.

The Database contains deal-related information, including the deal number, the DataStream Code of

the firm, which is used to match a firm’s accounting data from DataStream, the transaction value, the

shares percentage acquired by the bidder during the course of the transaction and share percentage

the bidder owned after the transaction, which will aid in determining whether a firm’s control has

been transferred in a transaction, the payment method, deal choice, deal type, deal attitude, SIC code

of the firm from which it can be judged whether or not the deal is diversified.

The sample involved 3,078 U.K. domestic acquisitions and 1,246 cross-border acquisitions into the

United Kingdom. Deals with a missing offer premium has been cleaned, yielding 1,826 acquisitions.

Observations with missing value of bidder 5-day CARs around the announcement date were excluded,

which left a sample of 1,435 acquisitions.34 Acquisitions with the information of payment method

were required, which left a sample of 1,212 acquisitions. Variables of both bidder and target firm

characteristics used in the regressions were not a missing value, which resulted in a final sample of

606 acquisitions, with 451 domestic acquisitions and 155 cross-border acquisitions.

34 The sample size reduces mostly because of bidders from countries without stock market returns, which is in line with

Erel et al. (2012).


109

Table 3.1 depicts the summary statistics for a sample of 606 public acquisitions studied in this chapter.

The mean value for the deal was $606.4 million. Of these 606 acquisitions, 380 were all-cash

acquisitions, 108 were all-stock acquisitions. There were 292 diversified acquisitions, 25 hostile

acquisitions, and 408 tender offers in the sample.

Figure 3.1 plots a time-series of number of deals for the full sample, the U.K. domestic sample, and

the cross-border sample. Overall, the sample period analysed covered the fifth and the sixth merger

waves, from 1993 to 1999, and started from 2003, respectively. The number of deals for the three

samples share a similar trend over the sample period. The number of deals increased throughout the

1990s, and peaked in 1999. They started to increase from 2003 after a sharp decrease in 2000, the

year the stock market crashed. The number of deals increased for several years from 2003 and 2006

before they declined in 2007.

Figure 3.2 plots the time-series of total deal value for the full sample, the U.K. domestic sample, and

the cross-border sample. The overall trend of the total deal value is consistent with that reported in

Figure 3.1. Interestingly, the overall trend of the total deal value of the cross-border acquisitions has

changed more dramatically than that of the domestic deals, and is larger than that of the domestic

sample in certain periods, such as 1999 to 2000, 2005 to 2007, and 2012 to 2013.

Table 3.2 shows the distribution of bidder nations. The sample consists of bidders from 13 countries.

Amongst them, a large majority are U.K. bidders, taking up 74.42% of the full sample. Of 451 cross-

border acquisitions, 67 acquisitions involve U.S. bidders who dominate foreign bidders in the cross-

border acquisition sample, taking up 43.22%. These results may imply that U.S. firms who share the


110

same language and have a similar cultural background to the U.K. firms are more likely to undertake

an acquisition of a U.K. target.

The thesis defines the offer premium (the target 52-week high) as the logarithmic term difference

between the offer price (the target’s highest stock price over 335 calendar days ending 30 days prior

to the announcement date) and target stock price 30 days prior to the announcement date.35 The

measure of offer premiums reflects the target expected gains during M&As, a concept which has been

widely used in related literature (Betton et al., 2009, Eckbo, 2009).36

A histogram of the difference between the offer premium and the target 52-week high, as shown in

Figure 3.3. The x-axis is in the range of -500% to 500%, and the y-axis shows the density. The shape

of the histogram indicates the extent to which the target 52-week high approaches the offer premium

within the sample.

The reference point effect on offer premiums was studied by controlling for a series of deal, bidder,

and target characteristics, as reported in Table 3.3. Edmister and Walkling (1985) documented that

cash payments are associated with higher offer premiums compared with stock payments, since

targets require higher offer premiums to compensate for the tax expense generated by capital gains.

Faccio and Masulis (2005) suggest that diversified mergers lead to greater uncertainty about bidder

35 Offer Premiums = log (Offer Pricei,t) – log (Stock Pricei,t-30). Target 52-week high = log (Target 52-week Highest Stock

Pricei,t-30) – log (Stock Pricei,t-30). The next trading day’s stock price was used when an offer was announced at a weekend.

The logarithmic term was used to counter the positive skewness bias of offer prices.

36 Prior research measures offer premiums with target abnormal returns (Schwert, 1996) questioned by Eckbo (2009). The

results are robust when a wide range of measures used for offer premiums, including target abnormal returns similar to

Schwert’s approach and the target’s price in a week and 3 months before the announcement date.


111

value, which implies that targets should be reluctant to accept risky offers unless higher offer

premiums are offered in compensation. A positive relationship between offer premiums and hostile

acquisitions is expected, since bidder managers pay higher offer premiums to persuade target

managers to accept an offer that may risk their professional career following a takeover. Bradley et

al. (1988) suggest that tender offers are associated with higher offer premiums since payment for the

manager goes to the shareholder.

Relative size is the deal value divided by the bidder market value. The market value (MV) is defined

as the current share price multiplied by the number of ordinary shares, expressed in the logarithmic

form. The market-to-book value (MTBV) is defined as the market value of the ordinary (common)

equity divided by the balance sheet of ordinary (common) equity in the company. Target volatility is

the standard deviation of target daily returns over the 335 calendar days ending 30 days prior to the

announcement. Run-ups are the pre-bid run-up prices calculated from 365 calendar days prior to the

takeover announcement date to seven calendar days before the takeover announcement date [-365, -

7]. All continuous data were winsorised at 1% and 99% levels to eliminate the outlier effect that both

extremely small or larger figures bias our results.

Relative size is defined as deal value divided by bidder’s market value, which measures the deal scale.

Previous research has highlighted a significant impact of the size effect on the offer premium (Asquith

et al., 1983; Dong et al., 2006). Firm size was measured as logarithmic term of market value of firms.

The mean value for firm size in the sample is larger for bidders than targets, 6.588 to 4.894, suggesting

that takeover bidders are generally stronger than their targets. This result is consistent with prior

M&A literature (Fuller et al., 2002; Moeller et al., 2004). A positive relationship between the offer


112

premium and bidder size, and a negative relationship between the offer premium and target size are

expected. According to Moeller et al. (2004), larger bidders tend to pay more for targets as they are

concerned with fewer restrictions in utilising the firm’s resources and thus become overconfident.

Alexandridis et al. (2013) suggest a robust negative relationship between the offer premium and target

size, implying that bidders’ acquisitions of a larger target are followed by a more complex process

for synergy creation, leading bidders to pay for targets with lower offer premiums in exchange for

expected synergies.

Firm’s growth opportunities were measured with market-to-book value (MTBV), which is consistent

with Rau and Vermaelen (1998). A higher MTBV indicates that the firm has a better investment

opportunity, whilst a firm with a lower MTBV suggests that the firm is short of investment

opportunities. Mean (median) bidder market-to-book value for the sample is on average higher than

that of the target, which suggests acquisitions involve a higher growth opportunities bidder and a

lower growth opportunities target. This is also predicted by the Q hypothesis of M&As (Lang, Stulz,

and Walkling 1989). A higher Q bidder is more likely to create synergies through acquiring a lower

Q target, thus it is willing to pay higher offer premiums for the targets. It is expected that offer

premiums are positively correlated to the bidder MTBV and are negatively correlated to the target

MTBV. A similar prediction can be made when the MTBV is a proxy for the firm’s valuation as per

Dong et al.’s work (2006). Thus, a high MTBV firm indicates that the firm is likely to be perceived

as a more overvalued firm whilst a low MTBV firm suggests the firm is likely to be perceived as

undervalued. It should be expected that a higher MTBV bidder in an attempt to dilute their

overvaluation tend to offer higher offer premiums to acquire a firm that is potentially being


113

undervalued (i.e. the lower MTBV targets). Therefore, in both cases, the offer premium is positively

(negatively) related to the bidder (the target) MTBV.

Pre-takeover price run-ups of both the bidder and the target (RunUps) were also controlled for while

analysing the reference point effect on the offer premium. Schwert (1996) found that the target price

run-ups increase the takeover costs of the bidder, indicating a dollar increases in target price run-up

leads to an increase of 1.13 dollars for the offer premium. It is suggested that the price run-ups are a

result of the market-wide valuation, leading the researcher to control for both the bidder and target

price run-ups. A mean (median) price run-up for the bidder is higher than that for the target. In

addition, the chapter accounts for the target volatility (Volatility) using the standard deviation of

target daily returns for the 335 calendar days ending 30 days prior to the announcement date.

Volatility of the firm represents the information asymmetry of the firm. A firm with high information

asymmetry tends to signal to the market that the firm is associated with high risks. Therefore, it can

be expected that the bidder tends to pay a higher offer premium when the true position of the target

firm is hard to justify.

The reference point effect on bidder announcement returns has been investigated by controlling for

a set of deal and bidder characteristics, since these factors have significant impacts on bidder

announcement returns, which are documented in prior M&A literature (Travlos, 1987; Rau and

Vermaelen, 1998; Alexandridis et al., 2010). More specifically, the method of payment for

acquisition (i.e. whether a means of payment for finance an acquisition is purely financed by stocks,

or cash), deal relatedness (i.e. whether or not the two firms involved in an acquisition is in the same

industry), deal type (i.e. tender offer or merger) and deal attitude (i.e. whether the deal is unsolicited)


114

were also accounted for. As to bidder characteristics, bidder size, bidder growth opportunities and

bidder pre-takeover period price run-ups were also taken into consideration.

Table 3.4 reports the mean value of a list of variables for two samples studied in this chapter: the

cross-border and the domestic acquisitions. The mean offer premium paid by the cross-border bidder

is 34.7%, which is significantly higher than that paid by the domestic bidder (27.6%), with a mean

difference of 7.1% at 1% significance level. There is no significantly difference for bidder abnormal

returns calculated by either the market-adjusted model or the market model. The mean value for the

target 52-week high is 27.7% for the cross-border acquisition sample, which is 4.4% higher than for

domestic acquisitions, suggesting that the reference point effect is more pronounced in the cross-

border acquisition sample than in the domestic acquisition sample. The cross-border bidder is found

to be stronger than the domestic bidder, reflected in significantly larger MV and higher MTBV than

those of the cross-border acquisition sample, suggesting that cross-border bidders who are more

disadvantageous in terms of information than domestic bidders should be strong enough to overcome

information asymmetry. The statistics further indicate that the information asymmetry for the cross-

border bidders tends to be larger than for their domestic counterparts.

3.4.2. Methodology

3.4.2.1. Piecewise linear regression and local polynomial smooth procedure

A non-linear relation between the offer premium and the target 52-week high is expected, based on

the shape of the value function proposed by prospect theory. According to prospect theory, targets

have a reference point in mind when selling the firms. Target shareholders may feel loss about the

deal their managers make when they receive a price that is lower than their 52-week high. This loss-


115

aversion occurs when targets lose the control of the firm. They may argue that the firm can perform

at a level similar to their 52-week high if their managers had not accepted the offer. Therefore, bidders

should take this into consideration and offer a price that is similar to the target 52-week high to

compensate the target’s mental feeling about the loss of control of the firm.

The OLS estimator was not applied in this context as it is a linear model assuming that the offer

premiums consistently increase with the target 52-week high. However, it is not the case as bidders

may walk away if the target’s current price is hard to capture its 52-week high price. Therefore, the

piecewise linear regression was employed to examine the relationship between the offer premium and

the target 52-week high. The piecewise linear function is a function of different linear segments, and

does not require an assumption of the shape of the data, which facilitates the analysis. It is worth

noting that bidders may not be able to offer a price to the target, whose current price is unlikely to

reflect their 52-week high. Bidders who pay according to such a 52-week high may find it hard to

rationalise the M&A motive to their shareholders. In addition, the marginal pain of the target

decreases when mental loss increases, leading additional costs according to the target reference point

to be reductant. Therefore, bidders may push up a price to acquire a target with a price closer to the

reference price and may not be willing to pay higher offer premiums for a target with a significantly

lower current price than the reference price. This suggests a consideration of diminishing sensitivities,

which is one implication of the prospect theory, while building the model to study the reference point

effect on the offer premium. The value function suggests that in both loss and gain domains, marginal

considerations tend to diminish relative to a reference point.


116

Though the quadratic function has been widely used in estimating the non-linear relationship, Gould

(1993) argue that the quadratic estimator has a reverse effect, which makes the slope hard to be

determined. The author suggests that the piecewise linear function outperforms both the linear

function and quadratic function in fitting the true shape of observations. Morck et al. (1988)

conducted a piecewise linear function to examine the relationship between management ownership

and the firm’s market value.

The local polynomial estimator will be employed, such a technique is to take the average of several

neighbourhood points and smooths the chart. Avery (2013) suggests that a local polynomial estimator

is the best linear smoother for point estimation. As depicted in Figure 3.4, the offer premium increases

with the target 52-week high, and this relationship can be observed clearly in three linear segments

with different slopes. The lower linear segment lies below 30% of the target 52-week high, the middle

linear segment is in the range of 30% and 70% of the target 52-week high, and the upper linear

segment is above 70% of the target 52-week high. 37 The piecewise linear function is presented in the

following, where the coefficient is the change in slope from the preceding group.

, 30(1) min 52 ,0.3i tf weekhigh (3.1)

, 30(2) max 0,min 52 0.3,0.4i tf weekhigh (3.2)

, 30(3) max 52 0.7,0i tf weekhigh (3.3)

37 The F-test and the Chow test were performed to examine the break points for the three target 52-week high segments

and the results show that the three segments are significantly different.


117

, 1 2 3 ,(1) (2) (3)i t i tOfferPremiums f f f (3.4)

where , 3052 i tweekhigh is defined as the logarithmic term difference between the target’s highest

stock price over 335 calendar days ending 30 days prior to the announcement date and the target stock

price 30 days prior to the announcement date. f (1) through f (3) are the transformation of the target

52-week high, the sum of f (1) through f (3) is equal to the target 52-week high. 1 ,

2 and 3 are

the coefficients for each linear segment, which represent the slopes of each segment.

3.4.2.2. Short-term event study

Short-term event study is a statistical tool for empirical research, and captures the market reaction

around the announcement date of M&As. Franks and Harris (1989) suggest that changes in stock

market should be fully captured in the several days around takeover announcement dates. Following

Brown and Warner (1985) who suggest that the standard procedures are well-specified and have

indicative power on firm performance, the daily stock returns with the short-term event study is

examined.

3.4.2.2.1. Market-adjusted model

Brown and Warner's standard short-term event study (1985) was used to calculate the CARs for the

bidder and the target in the five-day window (-2, 2) around the announcement date. The share returns

from the daily share price were calculated as follows:

, , , 1ln lni t i t i tR P P (3.5)


118

where ,i tR is the daily normal return of the firm i on day t, ,i tP denotes the share price of firm i on

day t, and , 1i tP denotes the firm’s share price lagged by a day.

, , , 1ln lnm t m t m tR P P (3.6)

where ,m tR relates to the daily normal return of the firm i while ,m tP and , 1m tP refer to the stock price

of the firm on day t and lagged by a day, t-1. The FTSE All-Share Index was used to proxy for the

value-weighted market index in the United Kingdom. The total market country index was used for

the other countries, which is defined as “TOTMK” plus “Country Code” in Datastream. For example,

US market index is “TOTMKUS”. The difference in the logarithm of the price was used performed

to capture the compound effect of the share price (Fama et al., 1969, Brown and Warner, 1985). The

abnormal returns was calculated with the market-adjusted model:

, , ,i t i t m tAR R R (3.7)

where ,i tR denotes the return of firm i on day t and ,m tR denotes the value-weighted market index

return of firm i on day t. The market-adjusted model assumes that α = 0 and β = 1 in our sample. The

abnormal returns (i.e. ARs) are summed to yield the CARs for the firm over the five-day window (-

2, 2) around the announcement date as follows:

, ,

0

n

i t i t

i

CAR AR

(3.8)


119

The market-adjusted model, as one of the main standard event methodologies, is widely employed to

measure firm takeover performance around the announcement date (Dong et al., 2006, Guo and

Petmezas, 2012). The model is believed to be particularly appropriate to assess the ARs of the firm,

as the model takes the market wide influence on the individual firm into consideration. We use

market-adjusted model to calculate CARs three and five days around the takeover announcement

date.38 The 5-day CARs are reported as our main results while the 3-day CARs are used as a

robustness test.

In order to assess the mean difference of CARs for the subsamples of domestic and cross-border

M&A deals, we employ t-statistics which is according to Seiler (2004). It allows the researcher to

investigate whether there are any distinguished characteristics of each portfolio CARs. The formula

is presented as follows:

( ) /

T

T

ARt

AR n (3.9)

where TAR denotes to the sample mean and as noted by Lyon et al. (1999:173), ( )TAR denotes to

the cross-sectional sample standard deviation for the sample of n firms.

3.4.2.2.2. Market Model

M&A literature also suggests abundant methods to be available to detect the short-term abnormal

returns. The market model was used as an alternative short-term event method to calculate the firms’

38 See also Fuller et al. (2002) and Faccio and Masulis (2005).


120

announcement returns, which is following Eckbo et al. (2016), who used the market model for a U.S.

study, and Gregory (1997), who used this method for a U.K. study. The model is specified as follow:

1it mt itR R (3.10)

where itR denotes holding period returns for firm i in the period t,

mtR denotes the value-weighted

market index. Similar to the market-adjust model, the FTSE All-Share Index was employed as a proxy

for the U.K. and total market country index was used for the other countries, which is defined as

“TOTMK” plus “Country Code” in DataStream, as described in the methodology session of market-

adjusted model. it denotes the error term. The market model parameters were estimated over the

window from 261 to 28 trading days prior to the announcement date [-261,-28], and calculate CARs

in both 3-day and 5-day event windows, and reported as robustness tests.

3.4.2.3. Multivariate analysis

The multivariate regression was used to examine the effect of offer premiums and the reference point

on firms’ short-run performance. According to Draper and Paudyal (2008), multivariate analysis is

superior in analysing the causation relation between the market reaction and related variables.

, , ,

1

N

i t i i t i t

i

CAR X

(3.11)

where ,i tCAR denotes the short-run CAR of firm i on day t, ,

1

N

i i t

i

X

denotes all the explanatory

variables in the model, and ,i t denotes the error term.


121

3.4.2.4. Two Stage Least Squares (2SLS)

A 2SLS estimator was constructed to test the effect of offer premiums on bidder performance around

the announcement date using the target 52-week high as an instrumental variable (IV). The use of IV

is to control endogenous variable. In this chapter, the regression of the offer premium on bidders

CARs was suspected with endogeneity issues. If the market views an offer price based on the

reference point as an overpayment, there are more negative market reactions generated by the 2SLS

estimator than the OLS estimator, and if the market views an offer price based on the reference point

as synergies, there are more positive market reactions generated the 2SLS estimator than the OLS

estimator. With the target 52-week high, the chain response of the offer premium to the bidder

announcement returns was able to examine. The use of the target 52-week high as an instrumental

variable is based on the theory of the offer premium. Larcker and Rusticus (2010), in their study

regarding the use of instrumental variables, suggest that instrument variables are eventually justified

by the theory. Equations of the two stages of the 2SLS estimator were presented as follows:

First stage:

(3.12)

Second Stage:

, 1 , , ,

1

N

i t i t i i t i t

i

CAR OfferPremiums x

(3.13)

where ,i tOfferPremiums in the first stage regression were calculated as the logarithmic term

difference between the offer price and target stock price 30 days prior to the announcement date.

,i tOfferPremiums , presented in the second stage, were the fitted value obtained from the first stage

, 1 , ,52i t i t i tOfferPremiums weekhigh


122

regression. ,52 i tweekhigh is the target reference point, which is defined as the logarithmic term

difference between the target’s highest stock price over 335 calendar days ending 30 days prior to the

announcement date and target’s stock price 30 days prior to the announcement date. The fitted value

from the first stage regression was obtained and used to replace the offer premiums in the second

stage regression. The target 52-week high is used as an instrumental variable of offer premiums as it

reflects the target performance, which is uncorrelated with bidder announcement returns and strongly

correlated with the offer premium.

The validity tests of the use of IV by performing Hausman test as well as the first-stage F-test were

conducted. The null hypothesis of Hausman test is that OLS has no measurement error or OLS

estimates are efficiency. The null hypothesis of the first-stage F-test suggests that there is no relation

between the endogenous variable and the instrumental variable (or there is no weak instrument

problem). According to Stock et al. (2002), the sampling distribution is generally nonnormal when

weak instrument problems exists, which lead to an unreliable estimation. Therefore, the first-stage F

test was employed to combat this problem, Stock et al. indicate the first-stage F-statistics should be

large enough to reject the null hypothesis.

3.5. Empirical results

3.5.1. The reference point effect on the offer premium

In this session, the piecewise linear regression of the offer premium on the reference point effect will

be conducted. Based on the prediction of prospect theory value function, our hypothesis proposes that

there is a positive correlation between offer premiums and the reference price in the U.K. market.

The main variable of interest to be investigated is the target 52-week high. Other variables that have


123

important impacts on the offer premium according to the M&A literature were controlled for in the

regressions.

The researcher first examines the correlation between the offer premium and the target 52-week high

with kernel density graph. As reported in Figure 3.3, the offer premium is highly correlated to the

target 52-week high. In most cases, offer premiums are paid around the target 52-week high,

suggesting that the reference point effect is pronounced in the sample.

Based on the diminishing sensitivity, it was expected that the offer premium tends to be more relevant

to the target 52-week high in the lower 52-week high level for the rationale that the target’s current

price is likely to reflect its recent best performance, leading the bidder to believe that it is possible to

generate synergies following an M&A deal, while the offer premium is less relevant to the target 52-

week high in the higher 52-week high level, implying that the bidder may walk away when the target’s

current price is unlikely to reflect its recent best performance.

Figure 3.4 supports our prediction by portraying a non-linear relationship between the target 52-week

high and offer premiums. The function predicts that targets have less marginal consideration about

loss. As noted earlier in the methodology section, the first linear segment lies below 30% of the target

52-week high (i.e. Low52-week high), the second linear segment lies between 30% and 70% of the

target 52-week high (i.e. Mid52-week high), the third linear segment lies above 70% of the target 52-

week high (i.e. High52-week high).


124

Table 3.5 shows a positive relationship between the offer premium and the target 52-week high, which

supports our hypothesis H1a. Specification (1) shows that a 10% increase of the target 52-week high

leads to an increase of 2.08% of the offer premium (coefficient = 0.208, t = 5.180). As we expected

earlier, the offer premium is less relevant to the target 52-week high when the level of the target 52-

week high is high or the current price is unrealistically far below the target 52-week high. Thus, it

leads us to study the reference point effect on the offer premium in the three different segments

predicted in the piecewise linear regression: the low target 52-week high, the mid target 52-week high

and the high target 52-week high. Specification (2) reports the target 52-week high on the offer

premium without any other control variables. It can be seen that the fitness of the specification is only

5.5%, which means the model might not be well-specified. Three different sets of variables in the

specifications (3) to (5) were taken into accounted. Specifically, the deal and target characteristics

were controlled in specification (3), deal and bidder characteristics were controlled for in

specification (4).

Specification (5) of this table shows the main results for the reference point effect on the offer

premiums. The offer premium is less relevant to the target 52-week high when the level of the target

52-week high increases. The offer premium increases by 4.33% for a 10% increase in the low level

of the target 52-week high, and increases by a 2.05% for a 10% increase in the middle level of the

target 52-week high, whereas the high level of the target 52-week high is insignificantly related to

the offer premium.39 Results obtained in this table suggest that bidders pay according to the target 52-

week high. However, they are rationally limited when the target 52-week high is unable to provide

any rationale motive for a higher price. In that case, bidders may find it hard to pay according to the

39 Our results are also consistent when controlling for the firm’s leverage and year-, industry-, and firm-fixed effects.


125

target 52-week high, which could create wealth for their shareholders when the target’s current price

deviates unrealistically far from the target 52-week high. The results obtained in this table can be also

interpreted within the prospect theory framework: while the price that deviates more from the

reference point tends to decrease the marginal considerations about the loss of the target (if the target

views selling the firm as a loss), leading higher M&A offer premiums are reductant for the bidder.

By doing so, bidders tend to be reluctant to pay offer premiums according to the target 52-week high

when the target 52-week high is high. The results are more pronounced compared with the U.S.

findings of Baker et al. (2012).

The results reconcile several predictions of prospect theory. People have reference-dependence bias

in that their decision on the offer premium is based on a salient piece of price information. Specifically,

the offer price is found to be set based on the target 52-week high. Bidders use the target’s recent best

performance as a reference price to suggest that the primary M&A motive for the bidder is synergy

generation, since targets whose reference point price is closer to their current performance is likely

to convince bidders to pay according to their reference prices.

The results are also in line with the diminishing sensitivity of prospect theory. It can be seen that the

lower level of the target 52-week high has more substantial impacts on the offer premium than the

higher levels. Bidder managers in the lower segment of the target 52-week high tend to believe

acquiring targets whose current price deviates greatly from the reference price is likely to leave more

room for synergy exploration and thus pay generous for targets than the scenario when the target’s

current price is closer to the reference price. While in the higher segments of the target 52-week high,

bidder managers tend to believe the target 52-week high is unable to deliver real support for the true


126

performance of the target, leaving the reference point effect on the offer premium less relevant. This

is in line Baker et al's view (2012) that bidders are cautious about those targets, whose current price

deviates greatly from the 52-week high, since they are more likely associated with bankruptcy risks,

violating the reference point effect. It is also possible that target managers are unable to bargain a

higher offer premium based on a high level of reference points.

Further, investors generally have a loss-aversion tendency, and such losses are gauged in relation to

the reference point. Based on these results, bidders pay a price related to the target 52-week high to

compensate for target shareholders’ mental loss when giving up the control of the firm. Target

shareholders, whose firm’s current price is close to the 52-week high, may presumably believe that

the firm would be better off if it was not acquired. In that case, bidders have to compensate the targets’

loss with an offer premium.

In addition to those explanations above in relation to prospect theory, it is expected that the stronger

legal environment and shareholder protection in the U.K. market than any other market may result in

a stronger reference point effect. One possible explanation is that in a market where the interests of

target managers and shareholders are closely aligned, managers are less likely to conduct any

frustrated actions against their shareholders. Thus, target managers are likely to reinforce their

bargaining power in M&As.40

40 The effect of corporate governance on the firm’s bargaining power is a recommendation for future research highlighted

in the Chapter 6 of this thesis.


127

It is worth noting that that the variable Cross Border is positively related to the offer premium,

reported in Table 3.5. Cross Border is a binary variable, taking value of 1 if acquisitions are

undertaken by cross-border bidders and 0 otherwise. Our results are in accord with the view of Rossi

and Volpin (2004) that stronger shareholder protection results in lower costs of capital favouring the

intense competition environment, increasing offer premiums. According to Edmister and Walkling

(1985), the offer premium represents the bargaining power of negotiators. In the U.K. settlement,

cross-border bidders have less bargaining power than domestic bidders engaging in acquisitions,

resulting in larger M&A offer premiums.

According to this prediction, the reference point effect will be further tested by cross-border

acquisition and domestic acquisition subsamples. It is proposed that there is a positive correlation

between offer premiums and the reference price when cross-border bidders acquire U.K. targets.

Targets with greater bargaining power are more likely to use the reference point to require a higher

offer premium.

Table 3.6 shows a slightly stronger reference point effect on the offer premium for the cross-border

acquisition subsample than for the domestic acquisition subsample. Several possible explanations for

this finding are offered. First, it is suggested that geographical difference creates greater information

barriers for cross-border bidders than for domestic bidders, which weakens cross-border bidders’

bargaining position. With less information about the target, cross-border bidders are likely to estimate

the target value with more relevant and easily obtainable information such as the target 52-week high.


128

Further, it can be argued that one of the main takeover motives for cross-border bidders is to access

the foreign capital market to integrate their resources, and obtain strong shareholder protection when

bidders from a country with relatively weaker shareholder protection than the target country. With

this in mind, cross-border bidders are likely to be influenced by the target 52-week high. According

to Alexandridis et al. (2013), acquiring large targets is a complex task which requires considerable

information, while it is impossible to assess the firm value given limited time and information. Thus,

valuing targets based on its target 52-week high simplify the acquisition process, giving cross-border

bidders quick access to the U.K. market. It is quite common that cross-border bidders are relatively

those large companies, as they should be strong enough to overcome barriers. According to Moeller

et al. (2004), managers from large companies have easier access to the firm’s resources, making them

more likely to overpay for targets.

Information disadvantages arising from geography weaken bidder bargaining position. Compared

with Baker et al.’s work (2012) focusing on the reference point effect on U.S. domestic M&A deals,

results obtained in this chapter show that the target 52-week high has greater predictability for offer

premiums for both cross-border and domestic acquisitions in the United Kingdom. Cross-border and

domestic bidders tend to have similar bargaining power, as reflected in the coefficients of the target

52-week high. However, the size effect distinguishes the two subsamples. Bidder size is more

significant for domestic acquisitions than for cross-border acquisitions, suggesting that domestic

bidders tend to be more consistent with the hubris hypothesis cited in Moeller et al.’s work (2004),

affirming that a larger firm is likely to be more overconfident than that of a smaller firm. The further

analysis aims to explore why cross-border bidders tend to pay similar M&A premiums relative to the

target 52-week high as domestic bidders.


129

In Table 3.6, the offer premium is positively related to the relative size, bidder size, the pre-bid target

run-up price and hostile mergers. It is also found that the offer premium is negatively related to target

size, target MTBV, target stock volatility, all-stock financed acquisitions and all-cash financed

acquisitions. These results are consistent with prior research (Schwert, 2000; Betton et al., 2008;

Eckbo, 2009). It is worth noting that the size of the bidder and target play an important role in offer

premiums. A larger bidder tends to pay generously for a smaller target, consistent with the findings

of Moeller et al. (2004) and Alexandridis et al. (2013). Contrary to the tax implication hypothesis, it

is suggested that the offer premium is negatively related to all-cash acquisitions, as reported in

specification (4). It can be interpreted as follows: cash is the greatest liquidity asset that can instantly

meet the target’s demands, thus the use of cash increases the bidder’s bargaining power, reducing

offer premiums.

To ensure that the model used here does not suffer from a muliticollinearity problem, i.e. the main

variable has enough explanatory power to the dependent variable, a test of variance inflation factor

(VIF) is employed. The test assumes that there is no correlation between the regressors when the VIF

value falls below 10. The presence of a multicollinearility problem in the model though does not

reduce the predictability of the model as a whole, however it increases the variance of the model,

making the model unstable, and reduces the accuracy of each individual variable included in the

regressions. The results obtained show little evidence of multicollinearity problems.41

41 It was found that the mean value for VIF is 1.75, and there is no VIF value beyond 4. The highest VIF value is 3.32,

which is the bidder MV, the second highest VIF value is 2.94, which is the target MV. The chapter reports the main results

of VIF test but omits the table for the sake of brevity.


130

3.5.2. The reference point effect on bidder announcement returns

This section examines the reference point effect on bidder announcement returns. A 2SLS regression

using the target 52-week high as an instrumental variable will be conducted. The rationale of using

the target 52-week high as an instrumental variable is because the offer premium is not a clean

measure, as it either represents the synergy or an overpayment. Hence, in order to discover the role

of the reference point effect on CARs, the 2SLS estimator is used. If the offer premium (driven by

the reference point effect) in the 2SLS leads to fewer announcement returns for the bidder than in the

OLS, the reference point plays a role of overpayment, otherwise, the reference point indicates the role

of synergies. It is proposed that bidder short-run performance is negatively correlated to offer

premiums when deals are made with regard to the reference price.

Table 3.7 shows that bidder announcement returns are either insignificantly or significantly

negatively related to the offer premium in different subsamples, as reported in specifications (1), (3),

and (5) respectively. It appeared that cross-border bidders pay higher offer premiums but outperform

domestic bidders. An acquisition of a U.K. target is good news to shareholders of cross-border bidders

who might believe that their rights to be protected in a more stringent legal environment, and thus

react positively to the bid announcement.

It is expected that the target 52-week high increases the offer premium, which further decreases the

bidder announcement returns. The Hausman test shows a p-value of 0.0004, suggesting that the 2SLS

estimator is more efficiency than the OLS estimator. The first-stage F-test shows a p-value of 0.000042,

indicating that the target 52-week high is a valid instrumental variable for the offer premium. A more

42 See Stock et al. (2002) for a survey of weak instrument problems.


131

negative relationship between bidder announcement returns and the offer premium became apparent

when using the target 52-week high as an instrumental variable compared with those without using

the target 52-week high as an instrumental variable, as shown in specifications (2) to (1), (4) to (3),

and (6) to (5) respectively. The market reacts more negatively to a bid announcement when bidders

pay based on the target reference price, suggesting that the reference price indicates an overpayment

in the U.K. M&As.43

These results are in line with prior M&A research. According to Rossi and Volpin (2004) and Moeller

and Schlingemann (2005), it is expected that the strength of shareholder protection in the U.K. would

weaken the bargaining power of cross-border bidders, thus using the target 52-week high is more

likely to receive shareholders’ consent. The sample was divided into cross-border and domestic

acquisition subsamples, as reported in specifications (2) and (3), and it was found that the target 52-

week high driven offer premiums were significantly negative related to announcement returns of

domestic bidders, and were not significantly related to those of cross-border bidders,44 suggesting

that shareholders of cross-border bidders acquiesce more easily with the target 52-week high as the

reference price. This chapter also provides additional support to the work of Kuipers et al. (2009),

who found that cross-border bidders in pursuit of stronger shareholder protection in the U.K. will pay

higher offer premiums as compensation. Further, the bidder size for cross-border acquisitions is more

negatively related to bidder announcement returns, which is consistent with the findings of Rau and

Vermaelen (1998) that small bidders tend to be more cautious than the large bidders. Taken together,

43 Similar conclusions can be drawn by using the market model in a window of 261 to 28 trading days prior to the takeover

announcement date. The results are reported in Table 3.9. 44 The obtained results of this chapter are robust to bidder announcement returns calculated with the market model.


132

cross-border bidders who have less bargaining power might have to pay higher premiums according

to the target 52-week high.

3.6. Robustness checks

3.6.1. The target 52-week high as a reference point

In this subsection, robustness results are provided for the target 52-week high as a reference point by

first employing other price measures related to the target 52-weeks: the target 52-week minimum and

the target 52-week mean. Moreover, the target monthly returns prior to the takeover announcement

date are utilised the expectation that the bidder may also extrapolate the past performance of the target.

Results on regressions of the offer premium on the reference point are robustness after checking for

different price measures related to the target 52-weeks. The target 52-week minimum and the target

52-week mean were taken into consideration besides the target 52-week high in terms of the reference

point price, as reported in Table 3.8. Huddart et al. (2009) provided an overall review with regard to

the reference price using both 52-week highs and 52-week lows. In a similar way, the target 52-week

mean is also considered as the reference price. It is generally believed that the target 52-week prices

tend to attract a great deal of investors’ attention, making more conservative investors likely to focus

on the firm’s mean value over 52 weeks prior to the announcement date since it reflects the firm’s

performance in a certain period of time rather than a snapshot price. The results show that the target

52-week high remains strong after checking for those proposed reference point price measures.

In addition, robustness checks were carried out by including the target past returns prior to the

takeover announcement date when studying the reference point effect on the offer premium (see Table


133

3.9). According to Rau and Vermaelen’s extrapolation hypothesis (1998), market investors tend to

forecast a firm’s prospects based on its past returns, creating a strong performance momentum around

a bid announcement for those with best past performance. Therefore, targets’ past returns of each

month were taken into consideration, reflecting the firm’s returns over the past months, which are

more likely be a proxy for the market perception of the firm’s valuation than a single item of price

information (i.e. the target 52-week high). The extrapolation effect on the offer premium was studied

by including the target past returns up to 12 months prior to the takeover announcement date. In

particular, specification (2) controls for the target monthly returns for the recent three months prior

to the takeover announcement date, specification (3) controls for the target monthly returns for the

recent six months prior to the takeover announcement date, and specification (4) controls for the

target monthly returns up to a year prior to the takeover announcement date. It was found that the

offer premium is positively related to the target 52-week high. Once again, our results show that the

target 52-week high is a proper measure for the reference point in an M&A deal.

3.6.2. The reference point effect on the offer premium by market conditions

In addition, the reference point effect on the offer premium by the market wide valuation was

examined. Investors’ have a rationale that bidders may under- or over-react to a target’s valuation

when the market valuation fluctuates. The sample periods are classified into High-, Low- and Neutral-

valuation markets according to Bouwman et al.'s method (2009). Results of Table 3.10 show that the

reference point effect is strong when the market-wide valuation is either high or low. It can be

interpreted that a high market valuation is likely to push the target’s current price closer to the 52-

week high, making bidders to believe that they can outperform the target’s recent performance. While

a low market valuation makes the bidder to believe that, the target 52-week high is a true indicator


134

for the target’s performance given that no other factors drive the firm’s value beyond or below its

fundamental value. Moreover, it was found that cross-border acquisitions tend to pay higher offer

premiums when the market valuation is high, implying that the high market wide valuation increases

information asymmetry and thus reinforces the target’s bargaining position over cross-border bidders.

3.6.3. The reference point effect on the offer premium across different subsamples

The reference point effect on the offer premium was studied by different nations with the expectation

that bidders whose nation was farthest away compared with that of the U.K. firms would tend to have

greater information barriers and be more likely to rely on the target 52-week high. In addition, the

reference point effect on the offer premiums in different M&A subsamples was analysed: the payment

method for the M&A deals, whether or not the deal is diversified and whether the deal involves a

negotiation between the boards of the two firms (i.e. tender offers and mergers).

A target’s bargaining position is likely to be reinforced when the distance involved between the two

firms’ home countries is large. The M&A sample was divided into two subsamples, the one includes

acquisitions comprising U.K. bidders, the other comprising acquisitions involving U.S. bidders,

farthest away compared with other cross-border bidders (see Table 3.11). It can be seen that the offer

premium increases by approximately 6% in the U.S. bidders compared with 4% in the U.K. domestic

bidders. Greater distance between the two nations impedes the transaction, making the bidder more

likely to look at the target reference point. The results obtained in this table are consistent with those


135

of Erel et al. (2012) and Uysal et al. (2008), finding that geographical distance plays a role in

information dissemination.45

The reference point effect on offer premiums across different subsamples by deal information was

also tested. The reference point effect on the offer premium being tested by the method of payment

(see Table 3.12), by whether or not the deal is diversified reported in (see Table 3.13), and the type

of merger reported in (see Table 3.14). Taken together, the results show that the target 52-week high

as a proxy for the reference point effect is very pronounced in different subsamples, and has a

substantial impact on the offer premium in different M&A subsamples.

3.6.4. The reference point effect on bidder announcement returns

Table 3.15 tabulates results on regressions of the offer premium on the bidder announcement returns

and with the target 52-week high as an instrumental variable of the offer premium. Instead of using

the market-adjust model to calculate the firm’s abnormal returns around the announcement date, the

market model is also used. The market model parameters were estimated based on an event window

from 261 to 28 trading days prior to the announcement date, i.e. [-261,-28]. The results obtained from

the market model are consistent with those by the market-adjusted model. More specifically, a 10%

increase in the reference point driven offer premium decreased the bidder announcement returns by

1.83% in the full M&A sample. While dividing it into two subsamples, the reference point effect leads

to more negative bidder announcement returns for domestic bidders than for cross-border bidders.

45 It can be argued that the information asymmetry are relatively low due to culture similarity between the two countries.

However, the sample included in this chapter is small, which is likely to result in bias in the results. Further investigation

by including more observations needs to be done to explain whether the culture or the distance between the countries

matter the bargaining power. We leave it to the future research, since the focus of this chapter is to test the target reference

point effect on the firm’s valuation.


136

In addition, bidder abnormal returns in a 3-day window around the takeover announcement date were

examined using both market-adjusted model and market model. The reference point effect on bidder

3-day announcement returns calculated with the market-adjusted model is reported in Table 3.16, and

results with the market model are reported in Table 3.17, which are consistent with the main results

reported in Table 3.7.

3.7. Conclusion

In this chapter, the reference point theory of M&As in the U.K. context has been examined. It was

found that the reference price measured by the target 52-week high is significantly positively related

to the offer premium. The reference point effect plays an important role in valuing the target firm.

Both the bidder and the target have reference-dependence bias, but the rationales behind using the

reference point need to be clarified in terms of different parties. Targets with greater bargaining power

are likely to use the reference point to ask for more offer premiums. It can be argued that bidders

paying according to the reference point in favour of targets is either because of the overconfidence or

because of information barriers. However, it is evident that the reference point effect is no longer

sensitive to bidders when the target’s current price deviated greatly from its reference price (or the

high level of the target 52-week high).

Further, the sample was extended to cross-border acquisitions and it was found that the reference

point effect is more pronounced to the cross-border acquisition subsample than to the domestic

acquisition subsample. It can be interpreted as follows: targets have greater bargaining power over

cross-border bidders than over domestic bidders, and they are likely to translate it into higher offer

premiums. Cross-border bidders may find it difficult in overcoming information barriers and might


137

have to pay an offer price based on the reference point. They might as well actively pay based on the

target 52-week high, the reason of being to reinforce their competitiveness to enter the United

Kingdom.

Finally, bidders paying based on the reference point receive negative market reactions, since

shareholders tend to believe that the managers are paying too much, thereby destroying their wealth.

The finding also suggests that bidder shareholders are pessimistic about their own profit-generating

ability. Together, results obtained in this chapter are consistent with the predictions of prospect theory.

Chapter 3. Tables & Figures

138

Table 3.1: Summary statistics for M&A sample

This table reports summary statistics for a sample of M&As announced between 1985 and 2014. The number N

is the number of deals per year. The third and fourth columns are the mean and the median of the deal value

respectively, reported in million U.S. dollars. The fifth and sixth columns present the method of payment

information. Number of stock only (number of cash only) is defined as the method of payment is 100% by stock

(cash). The number of diversified acquisitions, hostile acquisitions and tender offers are reported from the seventh

to ninth columns. “Diversification” refers to diversified deals in which the primary two Standard Industry

Classification codes are different between bidders and targets. “Hostile” refers to hostile bids. “Tender” refers to

tender offers.

Year N Deal Value ($mil)

Mean Median

No. of

Cash Only

No. of

Stock Only

No. of

Diversified

No. of

Hostile

No. of

Tender

1985 1 16.10 16.10 1 - 1 - -

1986 1 14.00 14.00 1 - 1 - -

1987 9 283.07 223.10 6 2 5 1 3

1988 10 135.52 72.36 8 - 5 - 4

1989 10 650.56 34.37 9 - 7 1 3

1990 3 225.56 115.07 2 - 1 - 1

1991 19 294.68 87.39 12 1 15 3 10

1992 8 248.12 31.89 4 2 7 1 4

1993 11 38.71 27.08 6 1 5 1 7

1994 21 160.01 42.18 15 3 8 2 10

1995 23 358.74 41.67 18 - 8 2 9

1996 20 644.40 133.14 10 6 14 1 11

1997 33 356.98 60.41 24 6 14 1 25

1998 56 655.03 78.95 34 14 21 1 40

1999 65 943.50 121.28 32 12 37 2 53

2000 58 533.48 153.60 37 12 28 3 47

2001 25 907.36 77.90 15 3 11 - 19

2002 20 946.02 24.98 14 4 3 - 19

2003 17 595.35 123.59 10 4 4 - 13

2004 22 461.28 120.23 12 7 13 - 16

2005 35 1037.71 330.44 20 7 19 2 25

2006 21 1082.69 210.80 12 4 10 2 19

2007 34 724.03 47.57 23 5 15 - 23

2008 13 376.26 55.36 10 1 7 1 11

2009 11 676.23 94.18 5 4 6 - 10

2010 26 315.46 22.95 19 3 15 - 9

2011 7 221.69 182.65 4 2 3 1 4

2012 13 444.14 64.08 10 2 4 - 6

2013 7 817.49 98.24 3 1 3 - 5

2014 7 570.73 83.14 4 2 2 - 2

Total 606 606.40 85.19 380 108 292 25 408


139

Table 3.2: Distribution of bidder nations

This table presents the number of acquisitions and percentage of the number of acquisitions by bidder nations.

The summary statistic are based on a sample of 606 public acquisitions toward U.K. public targets announced

between 1985 and 2014. Bidders are publicly listed firms in their domestic stock markets.

Bidder Nation Number of acquisitions Percentage

Australia 5 0.83

Canada 9 1.49

France 13 2.15

Germany 16 2.64

Hong Kong 5 0.83

Ireland-Rep 4 0.66

Italy 5 0.83

Japan 9 1.49

Netherlands 7 1.16

South Africa 7 1.16

Switzerland 8 1.32

United Kingdom 451 74.42

United States 67 11.06

Total 606 100.00


140

Table 3.3: Summary statistics for variables

This table presents the means, medians, and standard deviations of the variables. Panel A presents dependent

variables used in the regressions. Offer premiums are defined as the logarithmic term difference between the offer

price and the target stock price 30 days prior to the takeover announcement. Bidder 5-day CARs are bidder 5-day

announcement returns calculated by the market-adjusted model. Panel B presents independent variables, our main

variable the target 52-week high is defined as the logarithmic term difference between the target highest stock

price over 335 calendar days ending 30 days prior to the announcement and the target stock price 30 days prior to

the announcement. Deal characteristics, including the method of payment, deal attitude, type of deals, cross-border

deals and relative size. Here stock, cash, diversification, hostile, tender offer and cross-border are dummy variable,

taking value of 1 if acquisitions are 100% stocks, 100% cash, diversified merger, hostile merger, tender offer, and

cross-border acquisitions, and 0 otherwise. Relative size is the deal value divided by the bidder market value. The

market value (MV) is defined as the current share price multiplied by number of ordinary shares, expressed in

logarithmic form. The market-to-book value (MTBV) is defined as the market value of the ordinary (common)

equity divided by the balance sheet of ordinary (common) equity in the company. Target volatility is the standard

deviation of target daily returns over the 335 calendar days ending 30 days prior to the announcement. Run-ups

are the pre-bid run-up prices calculated from 365 calendar days prior to the takeover announcement date to seven

calendar days before takeover announcement date [-365, -7]. Accounting variables were collected one month prior

to the announcement date, and continues variables were winsorised at the 1% and 99% levels.

Panel A: Dependent Variables Mean Median Std Dev

Offer Premiums 0.294 0.263 0.206

Bidder 5-day CARs -0.002 -0.003 0.070

Panel B: Independent Variables

Target 52-week High 0.244 0.185 0.226

Deal Characteristics

Stock 0.178 - 0.383

Cash 0.627 - 0.484

Diversification 0.482 - 0.500

Hostile 0.041 - 0.199

Tender 0.673 - 0.469

Cross Border 0.256 - 0.437

Relative Size 0.353 0.150 0.549

Bidder Characteristics

Bidder ln(MV) 6.588 6.625 2.190

Bidder MTBV 2.490 1.760 4.617

Bidder RunUps 0.088 0.028 0.423

Target Characteristics

Target ln(MV) 4.894 4.760 1.846

Target MTBV 2.471 1.575 4.063

Target RunUps -0.013 0.005 0.382

Target Volatility 0.022 0.020 0.012

Number of Observations 606


141

Table 3.4: Summary statistics for variables by cross-border and domestic subsamples

This table presents the means of variables by U.K. domestic and cross-border acquisition sample. The first column shows the mean value of variables in cross-border acquisitions.

The second column shows the mean value of variables in U.K. domestic acquisitions. Differential is the mean difference between the cross-border and the U.K. domestic. The

t-statistics are reported in parentheses and calculated using the t-test and the Wilcoxon rank-sum test for the difference between the cross-border sample and the U.K. domestic

sample. Statistical significance at the 1% level, 5% level, and 10% level, denoted ***, ** and * respectively.

Cross-Border U.K. Differetial

(Cross-Border-U.K.)

Offer Premiums 0.347*** 0.276*** 0.071*** (19.492) (29.799) (3.743)

Bidder 5-day CARma -0.006 -0.001 -0.005 (-1.115) (-0.328) (-0.719)

Bidder 5-day CARmm -0.010* -0.003 -0.007 (-1.800) (-0.824) (-1.096)

Target 52-week High 0.277*** 0.233*** 0.044** (13.763) (22.783) (2.100)

Relative Size 0.211*** 0.401*** -0.190*** (8.531) (14.140) (-3.756)

Bidder ln(MV) 8.010*** 6.100*** 1.910*** (58.203) (60.994) (10.121)

Bidder MTBV 3.191*** 2.250*** 0.941** (13.000) (9.511) (2.197)

Bidder RunUps 0.089** 0.087*** 0.002 (2.474) (4.463) (0.048)

Target ln(MV) 5.538*** 4.673*** 0.865*** (39.692) (54.172) (5.135)

Target MTBV 2.761*** 2.372*** 0.389 (8.647) (12.307) (1.029)

Target RunUps 0.005 -0.020 0.025 (0.146) (-1.153) (0.696)

Target Volatility 0.025*** 0.022*** 0.003***

(24.188) (37.340) (2.883)

Number of Observations 155 451


142

Table 3.5: The piecewise linear regressions of offer premiums on the target 52-week high

, 1 , 30 2 , 30 3 , 30 ,min 52 ,0.3 max 0,min 52 0.3,0.4 max 52 0.7,0i t i t i t i t i tOfferPremiums wh wh wh

This table reports piecewise linear regression results for offer premiums on 52WeekHigh, controlling for various deal and firm characteristics. We document three individual

target reference point (i.e. 52WeekHigh) levels: Low52wh (0-30%), Mid52wh (30%-70%), High52wh (70% or above). Colum (1) reports the relation between the offer premium

and the target 52-week high, column (2) reports the relation between offer premiums and the three individual levels about the 52WeekHigh, column (3) controls for target and

deal characteristics, column (4) controls for the bidder and deal characteristics, column (5) controls for bidder, target and deal characteristics. Variable definitions are as in the

notes of Table 3.3. Robustness t-statistics are reported in parentheses. Statistical significance at the 1% level, 5% level, and 10% level, denoted ***, ** and * respectively, is

reported alongside the coefficients.

(1) (2) (3) (4) (5)

Offer premiums Coef. t-stat. Coef. t-stat. Coef. t-stat. Coef. t-stat. Coef. t-stat.

52wh 0.208*** (5.180)

Low52wh 0.322*** (3.618) 0.440*** (4.800) 0.326*** (3.692) 0.433*** (4.912)

Mid52wh 0.122 (1.057) 0.221** (1.985) 0.139 (1.217) 0.205* (1.807)

High52wh 0.150 (0.451) 0.178 (0.549) 0.001 (0.002) 0.089 (0.260)

Stock -0.071*** (-2.933) -0.063** (-2.476) -0.079*** (-3.380)

Cash -0.077*** (-3.461) -0.052** (-2.283) -0.075*** (-3.522)

Diversification 0.010 (0.669) 0.018 (1.179) -0.005 (-0.342)

Hostile 0.080* (1.781) 0.047 (1.028) 0.084** (2.048)

Tender 0.065*** (3.509) 0.070*** (3.812) 0.028 (1.486)

Cross Border 0.085*** (4.258) 0.049** (2.256) 0.046** (2.294)

Relative Size -0.030* (-1.831) -0.023 (-1.318) 0.045** (2.375)

Target ln(MV) -0.023*** (-4.785) -0.060*** (-7.869)

Target MTBV -0.002 (-1.275) -0.004** (-2.417)

Target RunUps 0.112*** (3.638) 0.087*** (2.704)

Target Volatility -2.553*** (-3.641) -2.926*** (-4.238)

Bidder ln(MV) 0.006 (1.361) 0.045*** (6.337)

Bidder MTBV 0.001 (0.513) 0.002 (1.179)

Bidder RunUps 0.044** (2.013) 0.025 (1.187)

Constant 0.244*** (21.468) 0.229*** (16.271) 0.375*** (7.906) 0.164*** (3.430) 0.284*** (5.986)

N 606 606 606 606 606

R2 0.052 0.055 0.185 0.138 0.260


143

Table 3.6: The piecewise linear regression of offer premiums on the target 52-week high of cross-border

and domestic acquisitions into the United Kingdom

This table reports the piecewise linear regression results for offer premiums on the target reference point of both

domestic and cross-border acquisitions into the U.K. market. It was documented three individual target reference

point levels: Low52wh (0-30%), Mid52wh (30%-70%), and High52wh (70% and above). Column (1) reports the

results for cross-border acquisitions. Column (2) reports the results for U.K. domestic acquisitions. Variable

definitions are as in the notes of Table 3.3. Equation is presented in the note of Table 3.5. Robustness t-statistics

are reported in parentheses. Statistical significance at the 1% level, 5% level, and 10% level, denoted ***, ** and

* respectively, is reported alongside the coefficients.

Offer Premiums (1) (2)

Cross-Border U.K. Domestic

Low52wk 0.480*** (2.779) 0.404*** (4.191)

Mid52wk 0.402 (1.600) 0.093 (0.768)

High52wk -0.149 (-0.226) 0.186 (0.469)

Stock -0.151** (-2.110) -0.072*** (-2.963)

Cash -0.063 (-1.241) -0.077*** (-3.285)

Diversification -0.010 (-0.310) -0.002 (-0.097)

Hostile -0.021 (-0.358) 0.157*** (2.748)

Tender 0.072* (1.931) 0.008 (0.377)

Relative Size 0.130* (1.682) 0.047** (2.451)

Target ln(MV) -0.044** (-2.202) -0.069*** (-8.700)

Target MTBV -0.004 (-1.402) -0.004** (-2.232)

Target RunUps 0.138** (2.518) 0.050 (1.370)

Target Volatility -3.049* (-1.967) -2.894*** (-3.576)

Bidder ln(MV) 0.024 (1.388) 0.054*** (7.266)

Bidder MTBV 0.015*** (2.651) 0.001 (0.666)

Bidder RunUps 0.023 (0.531) 0.024 (1.014)

Constant 0.317*** (2.701) 0.291*** (5.562)

N 155 451

R2 0.286 0.274


144

Table 3.7: The OLS regression of bidder announcement returns on offer premiums

First stage:

Second Stage:

, 1 , , ,

1

N

i t i t i i t i t

i

CAR OfferPremiums x

This table presents both OLS and 2SLS regression results of bidder market-adjusted model 5-day announcement

returns (CAR5) on the offer premium. The target 52-week high is used as an instrumental variable of the offer

premium. We report the regressions for both cross-border and domestic acquisitions into the United Kingdom.

Hausman test and first-stage F-test are reported below the table. Variable definitions are as in the notes of Table

3.3. Details of the methodology were presented in the Methodology Section of this Chapter. Robustness t-statistics


* respectively. (1) (2) (3) (4) (5) (6)

CAR5 Full Sample Cross Border U.K. Domestic

OLS IV OLS IV OLS IV

Offer Premiums -0.019 -0.207*** 0.015 -0.134 -0.036** -0.258*** (-1.360) (-3.268) (0.587) (-1.454) (-2.095) (-2.634)

Stock 0.007 -0.003 0.003 -0.012 0.008 -0.005 (0.807) (-0.291) (0.106) (-0.430) (0.762) (-0.408)

Cash 0.026*** 0.017* 0.001 -0.007 0.030*** 0.016 (3.506) (1.892) (0.048) (-0.346) (3.509) (1.351)

Diversification -0.012** -0.008 0.003 0.005 -0.014** -0.009 (-2.151) (-1.225) (0.311) (0.431) (-2.144) (-1.119)

Hostile 0.003 0.017 -0.003 -0.008 0.008 0.038 (0.177) (1.010) (-0.148) (-0.368) (0.421) (1.436)

Tender 0.002 0.016* -0.001 0.015 0.003 0.016 (0.345) (1.864) (-0.047) (0.935) (0.465) (1.542)

Relative Size -0.012** -0.019*** 0.017 0.019 -0.014** -0.022*** (-2.142) (-2.667) (0.907) (0.905) (-2.230) (-2.666)

Bidder ln(MV) -0.005*** -0.004*** -0.008** -0.010** -0.004** -0.003 (-3.848) (-2.629) (-2.324) (-2.540) (-2.508) (-1.577)

Bidder MTBV 0.001 0.001 0.000 0.002 0.001 0.001 (0.962) (1.297) (0.170) (1.013) (0.938) (0.820)

Bidder RunUps -0.015** -0.008 -0.017 -0.012 -0.016** -0.010 (-2.173) (-0.971) (-1.341) (-0.827) (-2.007) (-1.043)

Constant 0.029** 0.074*** 0.046 0.106* 0.025 0.082*** (2.128) (3.465) (1.199) (1.903) (1.635) (2.691)

N 606 606 155 155 451 451

R2 0.079 . 0.074 . 0.105 .

Hausman test X2 = 12.57

(p = 0.0004)

First-stage F test X2 = 13.23

(p = 0.0000)

, 1 , ,52i t i t i tOfferPremiums weekhigh


145

Table 3.8: Robustness test: Controlling for other relevant price measures

This table presents the piecewise regression results for offer premiums on the target 52-week high, controlling for

other price measures that might be used as the reference point prices. Column (1) reports the relation between

offer premiums and the target 52-week high. Column (2) reports the results by controlling for the target 52-week

minimum (52wMin), which is the target lowest stock prices over the recent 52 weeks, which is measured as

logarithmic term difference between target lowest stock price over 335 calendar days ending 30 days prior to the

announcement date and the target stock price 30 days prior to the announcement date, column (3) reports the

results by controlling for the target 52-week mean (52wMean), which with a similar measure to the target 52-

week high and 52-week minimum. Robustness t-statistics are reported in parentheses. Statistical significance at

the 1% level, 5% level, and 10% level, denoted ***, ** and * respectively.

Offer Premiums Other Price Measures

(1) (2) (3)

Low52wh 0.322*** 0.313*** 0.265**

(3.618) (3.224) (2.394)

Mid52wh 0.122 0.122 0.092

(1.057) (1.053) (0.766)

High52wh 0.150 0.150 0.109

(0.451) (0.448) (0.326)

52wMin 0.010

(0.310)

52wMean 0.055

(0.890)

Constant 0.229*** 0.234*** 0.243***

(16.271) (10.507) (11.747)

N 606 606 606

R2 0.055 0.056 0.057


146

Table 3.9: Robustness test: Controlling for target past returns

This table presents the piecewise regression results for offer premiums on the target 52-week high, controlling for

other price measures that used as the reference points. Column (1) reports the regression results by controlling for

segments of the target 52-week high, which is the target highest stock price over 335 calendar days ending 30

days prior to the announcement date and the target stock price 30 days prior to the announcement date, column

(2) reports the results by controlling for the target past returns in one, two and three months prior to the takeover

announcement date, column (3) controls for the target past returns up to 6 months prior to the takeover

announcement date, column (4) controls for the target past returns up to one year prior to the takeover

announcement date. Robustness t-statistics are reported in parentheses. Statistical significance at the 1% level, 5%

level, and 10% level, denoted ***, ** and * respectively.

Offer Premiums (1) (2) (3) (4)

Low52wk 0.322*** 0.325*** 0.336*** 0.298**

(3.618) (3.413) (3.376) (2.388)

Mid52wk 0.122 0.104 0.115 0.114

(1.057) (0.868) (0.941) (0.742)

High52wk 0.150 0.150 0.164 -0.206

(0.451) (0.452) (0.428) (-0.392)

Target return t-1 0.038 0.034 0.005

(0.540) (0.467) (0.065)

Target return t-2 -0.078 -0.065 -0.108

(-1.012) (-0.803) (-1.101)

Target return t-3 0.036 0.016 0.027

(0.509) (0.217) (0.286)

Target return t-4 0.103 0.094

(1.288) (0.884)

Target return t-5 -0.070 -0.055

(-0.813) (-0.486)

Target return t-6 -0.111 -0.082

(-1.341) (-0.710)

Target return t-7 -0.103

(-1.040)


(-1.422)


(-1.314)

Target return t-10 0.091

(1.050)


(-1.309)


(-1.355)

Constant 0.229*** 0.229*** 0.229*** 0.236***

(16.271) (14.332) (13.818) (10.927)

N 606 602 573 418

R2 0.055 0.059 0.066 0.070


147

Table 3.10: Robustness test: Reference point effect on offer premiums by market-wide valuation

This table presents the piecewise regression results for offer premiums on the target 52-week high by market

valuation. High Market is a dummy variable, taking a value of 1 if takeover months in the top 25% above past 5-

year average de-trended P/E of the UK market or market valuation is high, 0 otherwise. Specifications (1)-(3)

report the results when market-wide valuation is high (i.e. High), low (i.e. Low) and neutral (i.e. Neutral)

respectively. Details of this methodology were presented in methodology section of the Chapter 4. Other variable

definitions are as in the notes of Table 3.3. Equation is presented in the note of Table 3.5. Robustness t-statistics


* respectively, is reported alongside the coefficients.

(1) (2) (3)

High Low Neutral

Offer premiums Coef. t-stat. Coef. t-stat. Coef. t-stat.

Low52wh 0.494*** (4.322) 0.761*** (2.704) 0.119 (0.725)

Mid52wh 0.181 (1.156) 0.136 (0.517) 0.431* (1.709)

High52wh 0.297 (0.683) -0.245 (-0.365) -0.669 (-0.818)

Stock -0.042 (-1.417) -0.066 (-1.063) -0.164*** (-3.354)

Cash -0.021 (-0.784) -0.079 (-1.492) -0.162*** (-3.843)

Diversification 0.005 (0.235) 0.035 (0.951) -0.053* (-1.772)

Hostile 0.065 (1.339) -0.032 (-0.308) 0.112* (1.807)

Tender 0.029 (1.136) 0.045 (1.041) 0.017 (0.465)

Cross Border 0.049* (1.954) 0.068 (1.410) 0.000 (0.004)

Relative Size 0.080*** (3.260) 0.058 (1.013) 0.030 (1.199)

Target ln(MV) -0.073*** (-7.514) -0.038** (-2.063) -0.070*** (-4.785)

Target MTBV -0.006** (-2.589) 0.004 (0.856) -0.006 (-1.183)

Target RunUps 0.098** (2.397) 0.071 (0.864) 0.071 (1.189)

Target Volatility -2.925*** (-3.026) -3.282*** (-2.988) -2.505 (-1.407)

Bidder ln(MV) 0.055*** (5.833) 0.030** (2.197) 0.060*** (4.093)

Bidder MTBV 0.000 (0.037) 0.001 (0.207) 0.006* (1.882)

Bidder RunUps 0.042 (1.587) -0.041 (-1.172) 0.041 (0.883)

Constant 0.227*** (4.113) 0.184* (1.820) 0.383*** (3.326)

N 300 110 196

R2 0.324 0.291 0.319


148

Table 3.11: Robustness test: U.K. and U.S. bidder subsamples

This table reports the piecewise linear regression results for offer premiums on 52WeekHigh of both U.K. and U.S.

bidders’ acquisitions of U.K. targets. We document three individual target 52WeekHigh levels: Low52wh (0-30%),

Mid52wh (30%-70%), and High52wh (70% and above). Column (1) reports the results for U.K. domestic

acquisitions. Column (2) reports the results for U.S. bidders’ acquisitions of U.K. targets. Variable definitions are

as in the notes of Table 3.3. Equation is presented in the note of Table 3.5. Robustness t-statistics are reported in

parentheses. Statistical significance at the 1% level, 5% level, and 10% level, denoted ***, ** and * respectively,

is reported alongside the coefficients.

(1) (2)

Offer premiums U.K. Domestic Cross-Border Subsample (U.S. bidders)

Coef. t-stat. Coef. t-stat.

Low52wk 0.404*** (4.191) 0.593** (2.488)

Mid52wk 0.093 (0.768) 0.376 (0.943)

High52wk 0.186 (0.469) -0.661 (-0.633)

Stock -0.072*** (-2.963) -0.296** (-2.547)

Cash -0.077*** (-3.285) -0.096 (-0.861)


Hostile 0.157*** (2.748) -0.149 (-1.446)

Tender 0.008 (0.377) 0.141*** (2.773)

Relative Size 0.047** (2.451) 0.242 (1.158)

Target ln(MV) -0.069*** (-8.700) -0.053* (-1.976)

Target MTBV -0.004** (-2.232) -0.009 (-0.986)

Target RunUps 0.050 (1.370) 0.166 (1.585)


Bidder ln(MV) 0.054*** (7.266) 0.047 (1.442)

Bidder MTBV 0.001 (0.666) 0.025** (2.141)

Bidder RunUps 0.024 (1.014) 0.003 (0.046)

Constant 0.291*** (5.562) 0.239 (1.045)

N 451 67

R2 0.274 0.453


149

Table 3.12: Robustness test: Reference point effect on offer premiums by the method of payment

This table presents the piecewise linear regression results of offer premiums on the target 52-week high in different

payment method subsamples. Column (1) includes acquisitions that are purely financed by Cash, column (2)

includes those that are purely financed by stocks and column (3) includes those financed by a mixture of cash and

stocks. Other control variable definitions are as in the notes of Table 3.3. Equation is presented in the note of Table

3.5. Robustness t-statistics are reported in parentheses. Statistical significance at the 1% level, 5% level, and 10%

level, denoted ***, ** and * respectively, is reported alongside the coefficients.

(1) (2) (3)

Offer premiums Cash Stocks Mixed

Coef. t-stat. Coef. t-stat. Coef. t-stat.

Low52wh 0.420*** (3.557) 0.397** (2.138) 0.378* (1.774)

Mid52wh 0.248 (1.490) 0.150 (0.688) 0.348* (1.666)

High52wh 0.319 (0.661) 0.292 (0.432) -0.891 (-1.557)

Diversification 0.016 (0.786) -0.060* (-1.784) 0.015 (0.459)

Hostile -0.006 (-0.094) 0.278*** (3.813) 0.169*** (2.801)

Tender 0.055** (2.496) -0.053 (-1.188) -0.063 (-1.021)

Cross Border 0.045* (1.912) 0.016 (0.245) 0.022 (0.433)

Relative Size 0.056* (1.829) 0.061 (1.505) 0.102** (2.324)

Target ln(MV) -0.049*** (-5.431) -0.087*** (-3.812) -0.114*** (-4.953)

Target MTBV -0.005** (-2.428) 0.001 (0.106) -0.004 (-1.100)

Target RunUps 0.092** (2.329) 0.063 (0.889) 0.151** (2.042)

Target Volatility -3.632*** (-4.064) -1.707 (-1.318) -1.280 (-0.605)

Bidder ln(MV) 0.041*** (5.060) 0.052** (2.305) 0.100*** (4.415)

Bidder MTBV 0.004 (1.544) 0.000 (0.011) 0.002 (0.750)

Bidder RunUps 0.018 (0.582) 0.055 (1.486) -0.082 (-1.480)

Constant 0.172*** (3.332) 0.334*** (4.129) 0.222* (1.835)

N 380 108 118

R2 0.269 0.350 0.341


150

Table 3.13: Robustness test: Reference point effect on offer premiums for diversified and undiversified

M&A deals


payment method subsamples. Column (1) includes only diversified acquisitions (i.e. Diversificiation), column (2)

includes only undiversified acquisitions (i.e. Relatedness). Other control variable definitions are as in the notes of

Table 3.3. Equation is presented in the note of Table 3.5. Robustness t-statistics are reported in parentheses.

Statistical significance at the 1% level, 5% level, and 10% level, denoted ***, ** and * respectively, is reported

alongside the coefficients.

(1) (2)

Offer premiums Diversification Relatedness


Low52wh 0.468*** (3.515) 0.377*** (3.155)

Mid52wh 0.187 (1.261) 0.261 (1.551)

High52wh 0.168 (0.393) -0.115 (-0.213)

Stock -0.097*** (-3.032) -0.064* (-1.891)

Cash -0.061** (-2.003) -0.079** (-2.587)

Hostile 0.041 (1.028) 0.123 (1.592)

Tender 0.019 (0.727) 0.034 (1.309)

Cross Border 0.023 (0.843) 0.063** (2.155)

Relative Size 0.027 (1.039) 0.053* (1.961)

Target ln(MV) -0.046*** (-3.970) -0.071*** (-6.489)

Target MTBV -0.002 (-0.759) -0.007*** (-2.734)

Target RunUps 0.072 (1.581) 0.107** (2.393)


Bidder ln(MV) 0.037*** (3.662) 0.051*** (4.731)

Bidder MTBV 0.002 (0.837) 0.002 (0.700)

Bidder RunUps 0.008 (0.237) 0.035 (1.280)

Constant 0.264*** (4.323) 0.321*** (4.576)

N 292 314

R2 0.200 0.323


151

Table 3.14: Robustness test: Reference point effect on offer premiums for tender offers and mergers


payment method subsamples. Column (1) includes only tender offers (i.e. Tender), column (2) includes only

mergers (i.e.Merger). Other control variable definitions are as in the notes of Table 3.3. Equation is presented in

the note of Table 3.5. Robustness t-statistics are reported in parentheses. Statistical significance at the 1% level,

5% level, and 10% level, denoted ***, ** and * respectively, is reported alongside the coefficients.

(1) (2)

Offer premiums Tender Merger


Low52wh 0.410*** (4.167) 0.402** (2.389)

Mid52wh 0.218* (1.718) 0.151 (0.670)

High52wh -0.295 (-0.809) 0.846 (1.143)

Stock -0.086*** (-3.557) -0.122* (-1.693)

Cash -0.070*** (-3.307) -0.103 (-1.497)

Diversification -0.021 (-1.162) 0.035 (1.108)

Hostile 0.078* (1.793) 0.113*** (2.925)

Cross Border 0.059** (2.482) -0.000 (-0.005)

Relative Size 0.047** (2.296) 0.186** (2.393)

Target ln(MV) -0.070*** (-7.117) -0.053*** (-4.546)

Target MTBV -0.003 (-1.209) -0.006** (-2.479)

Target RunUps 0.051 (1.374) 0.147*** (2.710)

Target Volatility -3.034*** (-3.916) -2.193 (-1.478)

Bidder ln(MV) 0.051*** (5.689) 0.043*** (3.950)

Bidder MTBV 0.001 (0.763) 0.003 (0.683)

Bidder RunUps 0.025 (1.023) -0.007 (-0.179)

Constant 0.330*** (7.075) 0.250** (2.312)

N 408 198

R2 0.280 0.247


152

Table 3.15: Robustness test: Market model for 5-day CARs

This table presents both OLS and 2SLS regression results of bidder market model 5-day announcement returns

(CAR5mm) on the offer premium. We estimate market model parameters over the window from 261 to 28 trading

days prior to the announcement date [-261,-28]. The target 52-week high is used as an instrument. We report the

regressions for both cross-border and domestic acquisitions into the United Kingdom. Hausman test and first-

stage F-test are reported below the table. Variable definitions are as in the notes of Table 3.3. The 2SLS equation

is presented in the note of Table 3.7. Robustness t-statistics are reported in parentheses. Statistical significance at

the 1% level, 5% level, and 10% level, denoted ***, ** and * respectively. (1) (2) (3) (4) (5) (6)

CAR5mm Full Sample Cross-Border Bidders U.K. Bidders


Offer Premiums -0.016 -0.183*** 0.022 -0.152 -0.036** -0.210** (-1.155) (-2.982) (0.866) (-1.551) (-2.135) (-2.302)

Stock 0.005 -0.004 -0.003 -0.021 0.006 -0.004 (0.575) (-0.393) (-0.139) (-0.689) (0.597) (-0.352)

Cash 0.027*** 0.019** 0.008 -0.002 0.030*** 0.019* (3.573) (2.103) (0.399) (-0.089) (3.544) (1.707)

Diversification -0.011* -0.007 -0.002 0.000 -0.011* -0.007 (-1.950) (-1.145) (-0.162) (0.029) (-1.730) (-0.958)

Hostile 0.003 0.016 -0.003 -0.009 0.010 0.033 (0.207) (0.967) (-0.147) (-0.389) (0.500) (1.337)

Tender -0.001 0.011 -0.005 0.013 0.000 0.010 (-0.171) (1.343) (-0.421) (0.769) (0.036) (1.041)

Relative Size -0.012** -0.018*** 0.016 0.019 -0.014** -0.020*** (-2.118) (-2.618) (0.844) (0.838) (-2.283) (-2.657)

Bidder ln(MV) -0.006*** -0.005*** -0.009*** -0.012*** -0.004*** -0.004* (-4.267) (-3.130) (-2.625) (-2.793) (-2.708) (-1.945)

Bidder MTBV 0.001 0.001 0.001 0.004 0.001 0.001 (1.164) (1.457) (0.550) (1.383) (1.048) (0.958)

Bidder RunUps -0.040*** -0.033*** -0.036*** -0.030* -0.043*** -0.039*** (-5.934) (-4.321) (-2.709) (-1.924) (-5.527) (-4.274)

Constant 0.032** 0.072*** 0.050 0.120** 0.028* 0.072** (2.407) (3.501) (1.258) (2.030) (1.850) (2.554)

N 606 606 155 155 451 451

R2 0.134 . 0.123 . 0.161 .


(p=0.0016)

First-stage F test X2= 13.23

(p= 0.0000)


153

Table 3.16: Robustness test: Market-adjusted model for 3-day CARs

This table presents both OLS and 2SLS regression results of bidder market-adjusted model 3-day announcement

returns (CAR3ma) on the offer premium. The target 52-week high is used as an instrument. We report the


stage F-test are reported below the table. Variable definitions are as in the notes of Table 3.3. The 2SLS equation



(1) (2) (3) (4) (5) (6)

CAR3ma Full Sample Cross Border Bidders U.K. Bidders


Offer Premiums -0.017 -0.175*** 0.017 -0.105 -0.031** -0.220**

(-1.304) (-3.191) (0.899) (-1.332) (-2.131) (-2.578)

Stock 0.009 0.000 0.006 -0.006 0.009 -0.002

(0.993) (0.029) (0.211) (-0.248) (0.966) (-0.139)

Cash 0.028*** 0.020** 0.010 0.003 0.031*** 0.019*

(3.809) (2.557) (0.469) (0.171) (3.912) (1.811)

Diversification -0.010** -0.007 0.006 0.007 -0.014** -0.009

(-2.096) (-1.215) (0.666) (0.699) (-2.357) (-1.343)

Hostile 0.002 0.014 0.001 -0.003 0.006 0.031

(0.185) (0.968) (0.077) (-0.169) (0.396) (1.355)

Tender 0.004 0.015** -0.003 0.010 0.007 0.017*

(0.721) (2.110) (-0.233) (0.743) (1.069) (1.926)

Relative Size -0.008 -0.013** 0.012 0.013 -0.009 -0.015**

(-1.195) (-2.133) (0.493) (0.731) (-1.383) (-2.180)

Bidder ln(MV) -0.003** -0.002 -0.004* -0.006* -0.002 -0.001

(-2.424) (-1.526) (-1.720) (-1.841) (-1.338) (-0.693)

Bidder MTBV 0.000 0.001 0.000 0.002 0.000 0.000

(0.579) (1.002) (0.078) (0.927) (0.566) (0.532)

Bidder RunUps -0.013 -0.007 -0.016 -0.012 -0.013 -0.008

(-1.582) (-1.025) (-1.100) (-0.973) (-1.389) (-0.951)

Constant 0.009 0.047** 0.014 0.063 0.006 0.055**

(0.780) (2.541) (0.384) (1.316) (0.505) (2.063)

N 606 606 155 155 451 451

R2 0.075 . 0.055 . 0.102 .


(p=0.0083)


(p= 0.0003)


154

Table 3.17: Robustness test: Market model for 3-day CARs

This table presents both OLS and 2SLS regression results of bidder market model 3-day announcement returns

(CAR3mm) on the offer premium. We estimate market model parameters over the window from 261 to 28 trading

days prior to the announcement date [-261,-28]. The target 52-week high is used as an instrument. We report the


stage F test are reported below the table. Variable definitions are as in the notes of Table 3.3. The 2SLS equation



(1) (2) (3) (4) (5) (6)

CAR3mm Full Sample Cross Border Bidders U.K. Bidders


Offer Premiums -0.015 -0.151*** 0.026 -0.119 -0.033** -0.173**

(-1.133) (-2.848) (1.482) (-1.437) (-2.207) (-2.184)

Stock 0.007 -0.001 0.002 -0.013 0.007 -0.001

(0.768) (-0.066) (0.065) (-0.494) (0.766) (-0.073)

Cash 0.027*** 0.021*** 0.012 0.004 0.030*** 0.021**

(3.829) (2.731) (0.629) (0.231) (3.930) (2.233)

Diversification -0.010** -0.007 0.001 0.003 -0.012** -0.009

(-2.078) (-1.315) (0.129) (0.270) (-2.094) (-1.353)

Hostile 0.002 0.013 0.004 -0.001 0.005 0.024

(0.220) (0.899) (0.264) (-0.071) (0.348) (1.112)

Tender 0.002 0.011 -0.005 0.010 0.004 0.012

(0.315) (1.636) (-0.431) (0.723) (0.652) (1.422)

Relative Size -0.009 -0.013** 0.008 0.010 -0.010 -0.015**

(-1.355) (-2.297) (0.357) (0.538) (-1.542) (-2.257)

Bidder ln(MV) -0.004*** -0.003** -0.006** -0.008** -0.002 -0.002

(-3.042) (-2.150) (-2.101) (-2.229) (-1.584) (-1.056)

Bidder MTBV 0.001 0.001 0.001 0.003 0.000 0.000

(0.902) (1.283) (0.452) (1.432) (0.762) (0.745)

Bidder RunUps -0.026*** -0.021*** -0.026* -0.021 -0.027*** -0.024***

(-3.154) (-3.138) (-1.741) (-1.646) (-2.921) (-3.012)

Constant 0.014 0.047*** 0.018 0.077 0.011 0.047*

(1.285) (2.628) (0.515) (1.536) (0.866) (1.889)

N 606 606 155 155 451 451

R2 0.108 . 0.093 . 0.135 .


(p=0.0471)


(p= 0.0003)


155

Figure 3.1: Distribution of number of deals by bidders

This figure plots the time-series distribution of the number of deals for the full sample, the U.K. domestic sample,

and the CrossBorder sample. The U.K. domestic sample contains U.K. public bidders acquire U.K. public targets;

The Cross-Border sample contains those foreign public bidders acquire U.K. public targets. Y-axis indicates the

number of deals, X-axis represents year, spanning from 1985 to 2014.

0

10

20

30

40

50

60

70

198

5

198

6

198

7

198

8

198

9

199

0

199

1

199

2

199

3

199

4

199

5

199

6

199

7

199

8

199

9

200

0

200

1

200

2

200

3

200

4

200

5

200

6

200

7

200

8

200

9

201

0

201

1

201

2

201

3

201

4

Nu

mb

er o

f d

eals

All UK CrossBorder


156

Figure 3.2: Distribution of value of deals by bidder types This figure plots the time-series distribution of the total value of deals for the full sample, the U.K. domestic

sample, and the CrossBorder sample. The U.K. domestic sample contains U.K. public bidders acquire U.K. public

targets; The CrossBorder sample contains those foreign public bidders acquire U.K. public targets. Y-axis indicates

the value of deals in billion U.S. dollars, X-axis represents year, spanning from 1985 to 2014.

0

10

20

30

40

50

60

701

98

5

198

7

198

9

199

1

199

3

199

5

199

7

199

9

200

1

200

3

200

5

200

7

200

9

201

1

201

3

To

tal

va

lue

of

dea

ls i

n $

bil

All UK CrossBorder


157

Figure 3.3: Offer premiums relative to the target 52-week high

This figure presents the density of the difference between offer premiums and the target 52-week high, where

Offer Premiums is the logarithmic term difference between the offer price and the target stock price 30 days prior

to the takeover announcement and 52WeekHigh is the logarithmic term difference between the target’s highest

stock prices over the 335 calendars ending 30 days prior to the announcement date and the target stock price 30

days prior to the announcement date.

0.5

11.5

2

De

nsity

-5 5OfferPremiums - 52WeekHigh (%)


158

Figure 3.4: Nonlinear relation between offer premiums and the target 52-week high

This figure presents the nonlinear effect of the target 52-week high on offer premiums, where Offer Premiums is

the logarithmic term difference between the offer price and the target stock price 30 days prior to the takeover

announcement and 52WeekHigh is the logarithmic term difference between the target highest stock price over the

335 calendars ending 30 days prior to the announcement date and the target stock price 30 days prior to the

announcement date. We use local polynomial regression to smooth scatter plots, and set our sample where both

offer premiums and the target 52-week high are larger than 0, and less than 100%.

25

30

35

40

45

Offe

r P

rem

ium

s (

%)

0 10 20 30 40 50 60 70 80 90 10052 Week High (%)

159

CHAPTER 4: RELATIVE

REFERENCE PRICES AND M&A

MISVALUATIONS

Chapter 4. Relative Reference Prices and M&A Misvaluations

160

Abstract

This chapter examines the misvaluation hypothesis using a relative reference point (RRP) in the M&A

market. The reference point is the deviation of a firm’s current stock price from its 52-week high,

which reflects investors’ perception of the firm’s valuation. The market perceives the firm to be

overvalued when its current price is close to the 52-week high and to be undervalued when the current

price falls below its 52-week high significantly. The RRP indicates the extent to which the bidder is

more overvalued relative to the target. The sign of bidders’ overvaluation is revealed especially if

bidders pay takeover targets with stocks and with higher offer premiums. It was found that the RRP

accommodates the implications with respect to the misvaluation hypothesis. The results of this

chapter show that bidders prefer stock payments when the RRP increases, indicating that bidders are

overvalued and use their stocks to accelerate the process of overvaluation dilution. It became clear

that the RRP is positively related to the offer premium, suggesting that a relatively more overvalued

bidder tends to overpay for the target for the purpose of diluting overvaluation. With regard to the

relationship between the RRP and M&A outcomes, it became apparent that the RRP is found to be

positively related to target announcement returns, and is negatively related to the bidder

announcement returns. It also became apparent that the RRP is positively related to the stock bidders’

long-term abnormal returns. This result shows evidence of bidders’ rationality, eliminating concerns

about paying according to the RRP results in underperformance. Thus, the results are consistent with

the predictions of the misvaluation hypothesis and reference point theory.


161

4.1. Introduction

“Why could the typical investor expect any better success in trying to buy at low levels and sell at

high levels than in trying to forecast what the market is going to do? Because if he does the former

he acts only after the market has moved down into buying levels or up into selling levels. His role is

not that of a prophet but of a businessman seizing clearly evident investment opportunities. He is not

trying to be smarter than his fellow investors but simply trying to be less irrational than the mass of

speculators who insist on buying after the market advances and selling after it goes down. If the

market persists in behaving foolishly, all he seems to need is ordinary common sense in order to

exploit its foolishness.”

Benjamin Graham (1949: 31)

The misvaluation hypothesis explains an important motive of merger and acquisition (M&A)

activities. The theoretical model of Shleifer and Vishny (2003) predicts that the stock market drives

M&As. Overvalued bidders who serve the long-term interests of the shareholders will dilute

overvaluation through stock-financed acquisitions, as using stocks would accelerate such a dilution

process. The misvaluation hypothesis holds that bidders are rational whereas the market is irrational,46

which violates the efficiency market hypothesis and differs from what Roll’s (1986) hubris

management predicts. Following Shleifer and Vishny (2003), Dong et al. (2006) provided direct

evidence that bidders overpay for targets as long as bidders are overvalued relative to targets. Ang

and Cheng (2006), who investigated the long-term performance of stock bidders, found a positive

46 Shleifer and Vishny (2003) suggest that bidders are rational as they could time the market and use overvalued stocks

as a means of payment for acquisitions rather than holding them until they are corrected in the market. Bidders would pay

a price that is lower than the estimated synergy created by the combination of firms. The model assumes a target and a

bidder with stock volumes K and K1 respectively, price per unit Q and Q1 (where Q1>Q) and price per unit for the combined

equity S. As synergies can be achieved when S(K+K1)-(KQ+K1Q1)>0 the bidders would pay a price P to gain synergies

as long as it lies between Q and S, or Q<P<S.


162

relationship between overvalued stocks and long-term performance. Rhodes-Kropf and Viswanathan

(2004) provided a behavioural model suggesting that fully rational individuals make mistakes as they

tend to overestimate synergies, especially when market-wide valuation is high (Rhodes-Kropf et al.,

2005).

Conventional misvaluation measures face three major challenges. First, the measures relating to a

firm’s fundamental value cause estimation biases. Different firms could measure their assets in

different accounting approaches, such as adopting fair value and historical cost approaches vary

across the firms, which lead to variation of the firm’s fundamental value. Companies are also found

to manipulate accounting figures to raise the firm’s value, which is especially prevalent prior to the

financial crisis period. Secondly, existent misvaluation measures are mainly based on historical or

forward-looking information, such as price-to-book value (P/B), price-to-residual income value (P/V)

and earnings per share (EPS), which are less likely to reflect the latest status of the firm. Thirdly, the

frequently used ratio of MTBV is a proxy for both mispricing and investment opportunities of the

firm. According to Di Giuli (2013)47, firms with better investment opportunities should also increase

the practice of using stocks in acquisitions, leading to the same prediction as the misvaluation

hypothesis.

This chapter constructs a novel misvaluation measure: the relative reference point (RRP), which is

derived from the reference point effect48. Baker et al. (2012) defined the target reference point as the

47 Di Giuli (2013) proposed some post-merger investment-related proxies to disentangle the effects of mispricing and

investment opportunities.

48 Kahneman and Tversky (1979) proposed that with the reference point effect that people rely heavily on a single salient

piece of information while making decisions. Prior research has highlighted the 52-week high as a reference price ( Kliger

and Kudryavtsev, 2008; Barberis and Xiong, 2009).


163

deviation of a target’s current stock price from its 52-week high.49 Following this, the bidder reference

point in this chapter is defined as the deviation of a bidder’s current stock price from its 52-week high,

reflecting the extent to which the bidder is overvalued. It is argued that the firm’s 52-week high offers

the market an insight that the firm is overvalued, given that a firm’s 52-week high is the outcome of

a series of good news occurred in the past, the price will move toward to the firm’s fundamental value

once the price momentum disappears. When a firm’s current price is close to its highest stock price

over the year prior to takeover announcement, it suggests that the firm is still strong in price

momentum and should be reluctant to push up the price on the rationale that the firm is already

overvalued, whilst if a firm’s current price falls significantly below the 52-week high, substantiating

the market’s belief that the 52-week high is a sign of overvaluation. Based on this, the RRP is defined

as the difference between the target and the bidder reference points, indicating the extent to which the

bidder is relatively more overvalued than the target (i.e. the relatively more overvalued bidders).50

Baker et al. (2012) realised that the offer premium is positively related to the target reference point,

as the target uses its reference point to reinforce the bargaining position and demand high offer

premiums in an M&A deal, especially when the current price deviates significantly from its 52-week

high.51 On the other hand, bidders paying offer premiums based on the target 52-week high would

argue that they can outperform the target’s recent high and could generate synergies for the combined

firm following an M&A deal. In order to achieve the deal, bidders tend to pay for the target according

49 Their paper used’ the target reference point’ and ‘the target 52-week high’ interchangeably. In this chapter, the reference

point refers to the difference between the firm’s highest stock price over the period of 52 weeks prior to the M&A

announcement (52-week high) and the firm’s current price at the M&A announcement date. Such an approach makes this

reference point easier to interpret.

50 The construction of RRP is presented in the methodology section of this chapter.

51 They found a nonlinear relationship between the offer premium and the target 52-week high. In each segment, the offer

premium increases with the target 52-week high.


164

to its 52-week high. Their data support this view, finding that an offer price paid according to the

target 52-week high increase the deal success rate.

However, the idea of investigating the target 52-week high effect on the offer premium can explain

the negotiation process between the management teams of the two firms and how the firm is priced.

Baker et al. (2012) failed to reveal the bidders’ M&A motive as one cannot tell whether the proposed

offer price is a result of strangeness of the bargaining power of the target or any other considerations

of bidders. Their paper shows that bidders who believe that paying for acquisitions according to the

target 52-week high can generate wealth for shareholders. As a matter of fact, it destroys their wealth,

reflected in negative bidders’ announcement returns. Negative market reactions have been found

when bidders paying for acquisitions according to the target reference point results from overpayment.

Unlike those market investors, who are less experienced or lacking of information about the firms,

bidder managers can have more information about target’s value, and thus their M&A pricing

decisions should not be based on a single price number. It leads us to the question of whether bidders

have a similar view with the market with respect to the reference point. Specifically, the market might

believe that the target 52-week high represents a sign of poor performance when the target’s current

price falls below the 52-week high significantly, since it is less likely to rebound to the 52-week high,

whereas bidders paying for takeover targets according to the target reference point might believe that

the target is currently undervalued. It is evident that assessing a firm’s short-term performance does

not enable direct observation. Rather than short-selling the target’s stocks, bidders paying offer

premiums according to the target reference point have confidence in managing the deal afterwards.

Therefore, it can be argued that offer premiums based on the reference point are to guarantee

acceptance of the deal.


165

Another important question is whether the market also takes a look at the bidder reference point for

a similar reason to that for the target reference point, as this is also readily available to the public.

Simple intuition suggests that the market believes the bidder 52-week high is a highly relevant piece

of price information about the target value and treats it as a reference point, just like the target 52-

week high. The bidder reference point offers bidder managers an insight into how much they are able

to pay for the target. The bidder believes that the target 52-week high reflects the firm’s potential

profit-generating ability, and pays M&A offer premiums based on this, with the incentive of restoring

targets’ price discounts. On the other hand, target managers regard the bidder reference point as how

much in offer premiums they can possibly negotiate for when selling the firm. Baker et al. (2012)

found that targets demand high offer premiums based on the target 52-week high, relying on the

assumption that targets are not overvalued, otherwise they may be less likely to sell the firm for a

higher offer premium if they are undervalued. Therefore, the RRP reflects the market perception of a

firms’ valuation. The proxy explains relative valuations from the market angle, eliminating concerns

about any biases of a firm’s fundamental value.

Since both the target and the bidder reference point have a value in explaining how M&As are

structured, solely investigating the target reference point or the bidder reference point does not fully

account for the reference point effect on the M&A valuation. It should be expected that the market

will look at the bidder reference point believing that bidders whose current price is close to the 52-

week high are likely to create value given that they have performed well recently. This could also

reinforce the bargaining power of the bidders in an M&A deal, resulting in a lower offer premium

paid for the target, similar to Chira and Madura’s view (2015). In this situation, when bidders pay for

targets with a high offer premium, this would lead the market to believe that there is an overpayment.


166

However, according to the misvaluation hypothesis, managers could time the market by paying with

overvalued stocks for acquisitions. If the RRP is a suitable representation for relative valuations,

bidders looking at reference points of the two firms involved are actually going to time the market. If

this is the case, it remains largely unexplored why bidders rely on their reference point while paying

for the target, given better understanding about the firm.

The RRP is able to capture the market perception of the misvaluation of the two firms involved in an

M&A deal. Unlike Dong et al. (2006) and Rhodes-Kropf and Viswanathan (2004) who provide a

different view of bidder and target’s misvaluation,52 this researcher provides direct evidence that the

reference point should unify the investors’ view on the misvaluation of the two firms involved. George

and Hwang (2004) find misvaluation is driven by investors’ reactions. Lacking private information,

investors are reluctant to bid up a stock price when it is close to the 52-week high, based on the

rationale that it is previous good news that has driven a firm’s value beyond its fundamental value,

leading them to believe that a firm maybe overvalued. In contrast, investors should also be reluctant

to sell stocks of a firm whose price is far below the 52-week high, as this indicates that the firm maybe

undervalued. Extending their argument in the context of M&As, it becomes possible that managers

are theoretically highly committed in creating long-term value for the firm instead of enjoying short-

term profits from possible mispricing phenomena. Rather than Dong et al. (2006) who used the

difference between bidder’s and target’s P/B and P/V as proxies for the relative valuations between

the two firms, which are based on the firms’ fundamental value and argued that there is no actual

misvaluation but rather a managerial perception of misvaluation, the case of possible misvaluation

52 Dong et al. (2006) proposed price-to-book value and price-to-residual income value for bidder misvaluation whereas

Rhodes-Kropf and Viswanathan (2004) proposed a misvaluation measure based on an assumption that targets are rational,

and they overestimate merger synergies when valuation errors are sizable.


167

with the RRP from a market perspective is examined.

To the best of this researcher’s knowledge, this is the first paper to examine the role of the bidder

reference point in the context of M&A valuation. Following Baker et al. (2012), who proposed the

target reference point to explain how much bidders should pay for the acquisition target, it is

suggested that the bidder reference point would lead bidders to consider how much they are able to

pay for the acquisition target. Bidders who have performed poorly recently may find it difficult to

provide any rationale to pay the target with high offer premiums, while bidders whose stock price is

close to its 52-week high should be regarded as rich in financing resources, and thus dominate the

negotiation table, resulting in a low offer premium. It is much easier for them to persuade their

shareholders that the firm is stable and they can manage the deal well afterwards. In addition, in order

to boost a firm’s value and retain the market’s confidence in the firm, bidders should initiate an M&A

bid by sticking to the reference point, a way the market can judge, while remaining silent when the

firm’s price is close to the 52-week high tends to be very risky for the firm’s prospects. Barberis and

Xiong (2009) suggest that investors tend to sell stocks whose value has recently risen. Grinblatt and

Kelaharju (2001) and Huddart et al. (2009) report large abnormal sales’ volumes around the 52-week

high.

Another reason that bidders look at their own reference point is simply because bidder reference point

is an important component for managers to time the market. It is expected that managers can exploit

mispricing, as in the prior M&A literature that bidders are likely to pay with stocks for acquisitions

when they are overvalued (Ang and Cheng, 2006; Dong et al., 2006). However, the proxies used in

their papers for misvaluation do not accurately reflect the information such as a firm’s prospects and


168

managerial effectiveness, whilst the stock price is a relatively more comprehensive information.53

Therefore, it is suggested that the RRP, indicating the extent to which the market’s misperception of

the firm’s valuation, should provide similar findings as those predicted in the misvaluation hypothesis.

The RRP also signals to the target firm what price it could potentially negotiate with the bidder. If

targets believe, using the reference point, that they are overvalued, they would find it even harder to

justify this (overvaluation) given that they tend to be generally smaller and lack better investment

opportunities than bidders. This leads them to accept even more overvalued stocks for liquidity

purposes. In addition, an increase in the RRP also leads the target shareholders to believe that takeover

bidders are more attractive as it is more likely for the bidder to hit its reference price again more than

the target does. Hence, they might perceive that selling their firm to a well-run bidder would be more

likely to create value. According to Burch et al. (2012), targets tend to reserve bidders’ overvalued

stocks, maintaining that highly-valued bidders perform well or have better investment opportunities.

Therefore, stocks of the relatively more overvalued bidders are more attractive to targets.

Using M&As to investigate the relationship between the RRP and M&As is of great interest mainly

for two reasons. First, the RRP is a direct misvaluation measure that captures the market’s perception

of the firm’s valuation, avoiding biases caused by a firm’s fundamental characteristics (Lin et al.,

2011). M&As serving as a major corporate investment activity draw a great deal of investors’

attention. With limited information and limited time in which to process that information, investors

are likely to make decisions based on the most current market perceptions of a firm’s valuation,

53 The existence of mispricing in the short term has been confirmed even if the market is efficient, such as investors’

underreaction (Debondt and Thaler, 1985).


169

making the RRP a suitable valuation proxy for this testing ground. Secondly, the RRP will facilitate

the M&A process. Bidders can identify the relative overvaluation through the RRP, as an increase of

the RRP would potentially drive a relatively more overvalued bidder to dilute overvaluation through

acquisitions.

Analysing a sample of 1,878 U.S. domestic public acquisitions announced between 1985 and 2014

and using the RRP to test the predictions of the misvaluation hypothesis, it was found that the

propensity to use stocks as a means of payment for acquisitions increases with the RRP, which is

more pronounced when market-wide valuation is high. The results of this chapter are robust after

applying additional controls. Moreover, the relatively more overvalued bidders tend to pay higher

offer premiums. The results continue to hold after endogeneity checks that may become necessary

due to omitted variable biases. Finally, multivariate analysis results show that the RRP plays a role in

bidders’ overpayment in the short term. However, it became evident that higher offer premiums paid

by the relatively more overvalued bidders are translated into less negative abnormal returns in the

long term, suggesting that bidders who pay for acquisitions according to the RRP protect the wealth

of shareholders. Overall, the findings in this chapter are consistent with the predictions of the

misvaluation hypothesis and reference point theory.

This chapter makes three distinct contributions to the literature. Firstly, it explains the misvaluation

hypothesis from the perspective of the reference point. A dynamic valuation framework with the RRP

is developed, based on market perception of a firm’s valuation. The RRP overcomes the effect of

market-wide valuation for an individual firm, as the two firms involved experience the same market

conditions. A firm’s current price that is close to its 52-week highest point is more likely to be driven


170

by market-wide conditions (sentiment) and therefore associated with larger market valuation errors,

increasing the probability of the firm being overvalued. Using the RRP, it is possible to examine how

the valuation difference of the deviation of the two firms drives M&As. Hence, the bar is raised to

the market level, eliminating estimation biases arising from the use of a firm’s fundamental value.

Furthermore, the RRP is easily observable, which encourages investors to use it as a valuation

benchmark. The Results of This chapter suggest that the RRP is able to accommodate the implications

of the misvaluation hypothesis.

Secondly, the chapter provides direct evidence that more experienced investors behave similarly to

less experienced investors in major corporate investment decisions. However, the market and

managers may interpret the reference point effect differently as the market looks at the offer premium

paid according to the RRP as a result of overpayment by hubris-infected bidders, while bidders paying

for acquisitions according to the RRP is as a result of timing the market. The results indicate the less

negative long-term abnormal returns for the relatively more overvalued stock bidders compared with

the relatively less overvalued (or more undervalued) stock bidders.

Thirdly, this chapter offers a new insight into the method of payment hypothesis. It is suggested that

the sign of relative overvaluation is well identified by the RRP, which relaxes the assumptions of

irrational targets (Shleifer and Vishny, 2003) and valuation error misled targets (Rhodes-Kropf and

Viswanathan, 2004) related to the target’s motive of accepting overvalued stocks,54 as both bidders

and targets can identify any relative overvaluation in an M&A deal. Bidders paying stocks instead of

54 Common assumptions on whether the target will accept the overvalued stocks suggest that targets either have a cash-

out purpose (Shleifer and Vishny, 2003) or are misled by the market perception (Rhodes-Kropf and Viswanathan, 2004).


171

cash for a larger RRP acquisition reduce offer premiums while paying cash in a lower RRP acquisition

reduce offer premiums compared with the case of the higher RRP acquisitions. Hence, the RRP

justifies the method of payment choice.

The remainder of this chapter is organised as follows: Section 2 provides a range of literature that is

related to the misvaluation hypothesis and reference point theory of M&As. Section 3 designs

hypotheses with regard to the relations between RRP and the method of payment, offer premiums and

M&A outcomes in the short and long runs. Section 4 summarises the data and presents the

methodologies. Section 5 analyses the empirical results. Section 6 conducts further robustness checks

regarding the role of the RRP played in the M&A surveyed. Section 7 concludes the chapter.


4.2.1. The misvaluation hypothesis

Shleifer and Vishny (2003) developed a theoretical model of M&As based on the assumption that the

market is inefficient whereas managers are rational. With the assumption of the misvaluation

hypothesis, managers are able to explore the market’s mispricing given that firms are not fairly valued.

The authors suggest relative valuations of the two firms involved in an M&A deal as an important

factor explaining why M&As are initiated, the method of payment choice for financing M&As, and

the valuation consequences of the M&A wave. Their theoretical model predicts that bidders sacrifice

the short-term benefits in exchange for the firm’s long-term value. Overall, the model has raised two

important implications to be investigated: the method of payment choice and the long-term

performance of the combined firm.


172

The misvaluation hypothesis is related to the method of payment hypothesis. It suggests that the

method of payment choices would reflect bidders’ M&A motive. Bidders who offer targets with

stocks have long-term horizons, in that they aim to create value instead of reaping mispricing benefits

in the short run. Shleifer and Vishny (2003) suggest that the net long-term effect of stock deals would

generate short-term losses but long-term ‘gains’. Though stock-financed acquisitions bring forth

negative returns to the bidder in the long run, those acquisitions benefit bidders who pay with stocks

that were significantly overvalued prior to any acquisitions. As noted by Shleifer and Vishny (2003:

301), ‘returns are just not as negative as they would have been without the acquisition’, which

suggests that overvalued stocks used as a means of payment for acquisitions are treated as a cushion

for the collapse of the stocks in the long run. In this respect, a firm’s value is enhanced with

overvalued stocks even though there are negative long-term returns.

Moreover, the misvaluation hypothesis suggests that the method of payment clarifies the market

important information about a firm’s value. It is generally believed that bidders who pay targets with

stocks would signal to the market that their firms are overvalued or their firms are relatively more

overvalued than the targets, whilst targets paid with cash are perceived to be undervalued relative to

their fundamental value. This is also consistent with the signalling hypothesis of Myers and Majluf

(1984). The major difference between the two hypotheses regarding the method of payment is that

the signaling hypothesis assumes that investors are asymmetrically informed in that they do not have

private information about the other firm and pay with stocks to deal with uncertainty, whereas the

misvaluation hypothesis relaxes this assumption and posits that the primary motive of paying with

stocks is that managers try to time the market and aim to create value for the firm. Therefore, the

misvaluation hypothesis interprets the long-run reason for the bidder to finance an M&A deal,


173

reasoning that managers who pay with overvalued stocks try to time the market.

The theoretical model of misvaluation hypothesis has given rise to many important implications. First,

the issue on whether all-stock acquisitions are driven by misvaluation has raised doubts. Savor and

Lu (2009) imply that all stock acquisitions are driven by misvaluation, assuming that managers try to

time the market by paying for targets with overvalued stocks and highly valued bidders have the

greatest incentive to undertake an acquisition. Their study focuses on long-term performance of stock

bidders and finds that though stock acquisitions destroy the firm value, stock bidders of successful

deals outperform those who fail to complete an M&A deal. Analysing a sample of 12,578 U.S.

mergers between 1962 and 2000, the authors found that stock acquisitions generate long-term

negative abnormal returns for a firm. Despite this, the mean difference of buy-and-hold abnormal

returns between successful deals and unsuccessful deals increase from 13.6% to 31.2% over the three

years following an acquisition,55 suggesting that stock acquisitions are a result of managers timing

the market. In contrast, Fu et al. (2013) find evidence that not all stock acquisitions are driven by

misvaluation. Among 1,319 stock-financed acquisitions studied in their sample, one third is not

motivated by stock overvaluation, which is neither relative to its fundamental value nor relative to

the target valuation. Eckbo et al. (2016) documented more direct evidence contradicting the notion

that stock acquisitions are market-driven. Investigating a sample of 4,919 merger bids between 1980

and 2008, the authors found that bidders paying with stocks as a means of payment for acquisitions

tend to be small firms and with low leverage, implying that the reason for bidders to pay with stocks

is that the firm is financially constrained and limited in free cash flows. In addition, stock acquisitions

55 BHARs were calculated with the market-adjusted model. Their paper shows a similar trend when BHARs were

calculated with the calendar-time portfolio model.


174

are more likely to occur in high-tech industries and when target and bidder are in geographical

proximity, suggesting the reason for targets to accept stocks is because that they have information

about the true position of the bidder, eliminating the concern of information asymmetry.

The second concern relating to the misvaluation hypothesis is whether bidders paying with

overvalued stocks could create value for the firm. Shleifer and Vishny (2003) suggest that the long-

run reason for the bidders to use stocks is to create value for the firm as holding stocks in an

overvalued market will collapse the firm’s value when the market incorporates information and

downgrades the value accordingly. With this in mind, bidders are rational. Consistent with this

prediction, Ang and Cheng (2006) found that the long-term performance of stock bidders and

overvaluation are positively correlated. Analysing a sample of over 3,000 mergers between 1981 and

2001, the authors found significantly positive abnormal returns for stock bidders in the estimation

period of one day prior to the takeover announcement date and three years following merger

completion. The results continue to hold compared with those drawn from the control group that

contains similarly overvalued firms who do not undertake an acquisition. The result shows that

overvalued bidders outperform those similarly overvalued non-acquirers.

Lin et al. (2011) explain underperformance as long-run reversal of initial overvaluation rather than

an overpayment. Their quartile analysis results of buy-and-hold abnormal return by bidder’ P/V show

that overvalued bidders underperform their undervalued counterparts in the long run over three years

following an acquisition. Their results are robust in the situation where the offer premium and other

relevant variables were taken into consideration in the multivariate regression of BHARs on bidder

P/V. The rationale of controlling for the offer premium is that bidders with high P/V are also those


175

paying high M&A offer premiums, leading to the same prediction that overpayment results in

underperformance. In order to capture the net effect of the market correction on the firm overvaluation,

the authors used multivariate regressions to control for the offer premium after univariate analysis. In

addition, their data show that the average offer premium for the most overvalued bidders are

insignificantly different from the least overvalued bidders, 56.21% and 53.64% respectively, thus

weakening evidence that overpayment triggers underperformance in the long run.

Fu et al. (2012) argued that long-term underperformance is due to overpayment. They found that

overvalued firms tend to pay too much for acquisitions and attribute this to the weakness of the

corporate governance structure of a firm. The authors measure post-merger performance of the firm

with both the market-based approach as well as the accounting-based approach. Using the accounting-

based approach is based on the rationale that the initial overvaluation of the firm will be corrected

regardless of the occurrence of an acquisition. This may not be captured by conventional market-

based approaches. The authors thus assessed the firm’s long run performance with the firm’s ROA,

which is a proxy for firm’s operational performance rather than the market reaction, taking managerial

effectiveness into account. Overall, both approaches indicate negative long-term synergies for the

firms whose corporate governance structure is of inferior quality.

Since the long-term benefits of an acquisition are nearly zero as predicted by the misvaluation

model,56 leading to a question as to why target managers are willing to accept bidders’ overvalued

stocks. Shleifer and Vishny (2003) interpreted this as target’s short-horizons. With this in mind,

targets tend to voluntarily accept overvalued stocks to raise the liquidity of the firm or simply trade

56 Given the fact that value creation for the bidder, in return, damages the target in the long run.


176

them after acquisitions for a cash-out purpose. Another explanation is that this matter is attributed to

the agency problems of the target firms where managers are offered with lucrative benefits, such as

good pay or being retained in the top position of the bidder firm following acquisitions, leading them

to give up control of the firm.

However, the assumption of short-horizon targets has been questioned by a number of M&A studies.

Rhodes-Kropf and Viswanathan (2004) suggest a rational target model whereby rational targets

should not be willing to accept overvalued stocks, as they clearly know about the danger of holding

stocks in an overvalued market. The authors suggest that both target and bidder managers are well-

informed about their own firms but they lack private information about the other firm. Such private

information of the firm tells managers whether the other firm is undervalued or overvalued but cannot

tell managers what drives such misvaluation. A rational target is able to assess the synergies of the

merger and decide not to accept overvalued stocks so as to prevent the value of the firm from being

damaged. However, the view is distorted during the periods when market-wide valuation is high. As

it is explained by Rhodes-Kropf and Viswanathan (2004), ‘the rational target correctly filters on

average but underestimates the market wide effect when the market is overvalued and over estimates

the effect when the market is undervalued.’ Thus, an overestimation of synergy positively correlates

with valuation errors. Despite target managers being rational, they may find it hard to distinguish

synergies from misvaluation without knowing what drives misvaluation. Their data indicate that

stock-financed acquisitions are prevalent during high market-wide valuation periods when rational

targets tend to make mistakes by wrongly accepting bidders’ overvalued stocks.

Following this, Rhodes-Kropf et al. (2005) suggest that rational targets tend to accept overvalued


177

stocks when valuation errors mislead their estimation on synergies. The authors find the volume of

stock acquisitions and valuation errors are positively correlated. These authors use regression

techniques to break the market-to-book ratio (M/B), the conventional misvaluation measure, into

three individual components: the firm-specific misvaluation, the sector-wide misvaluation and the

long-run value to book which captures long-run growth opportunities. Their data suggest that the

firm-specific level and the sector-level are the most important factors driving misvaluation. More

specifically, bidders on average have higher firm-specific errors than targets, which is on average

60%. Misvaluation can reconcile about 15% of merger activity at the industry level. Rhodes-Kropf et

al.’s study (2005) shows that low long-run value-to-book firms buy high long-run value-to-book firms,

suggesting that low growth opportunities firms acquire high growth opportunities firms. Overall,

misvaluation-driven mergers are as a result of valuation errors, contradicting to what the misvaluation

hypothesis assumes.

Likewise, Di Guili (2013) finds rational targets, suggesting that target managers care about the quality

of mergers rather than private benefits. In this regard, stocks are associated with better investment

opportunities. So as to distinguish the effects of misvaluation and growth opportunities, the author

proposes a new proxy for investment opportunities of the firm and finds there is positive correlation

between stock payment and the firm’s prospects. Following this line of literature, Vermaelen and Xu

(2014) explained rational targets with capital structure theory. They found that rational target

managers are likely to accept overvalued stocks as they believe bidders paying with stocks will reach

an optimal capital structure. They proposed a prediction model regarding the method of payment,

which predicts that the bidder pays cash when the market expects cash payment. If bidders pay cash

when the model predicts stocks, the long-term abnormal returns will be positive based on the rationale


178

that the market under-reacts to the firm. Whilst bidders pay with stocks as the market predicted, their

long-term abnormal returns will be negative, assuming that the market can identify the motive for

financing an M&A is overvaluation. The prediction model explains 89% of Vermaelen and Xu’s

M&A sample (2014). It indicates that cash can generate positive long-term abnormal returns on the

condition that the market recommends stocks, whereas bidders should pay cash when the model

predicts cash, indicating that target managers are rational in that they would not accept overvalued

stocks without a reasonable explanation.

The misvaluation hypothesis explains who buys whom and the medium of payment. Dong et al. (2006)

conducted a thoroughly investigation into the relationship to these predictions. The authors analysed

pre-takeover valuation of targets and bidders by using the firms’ market price-to-book value of equity

(P/B), and market price-to-residual income (P/V). This latter misvaluation measure has been largely

ignored by previous studies. Their sample consists of 3,732 U.S. acquisitions announced between

1978 and 2000. Their study suggests that these two misvaluation measures are highly correlated but

P/B tends to have more explanatory power over the misvaluation hypothesis than P/V on the rationale

that managers are becoming more familiar with P/B than P/V. According to Dong et al. (2006), P

denotes the market value of the firm while both B and V denote the firm’s fundamental value. Thus,

the essence of these two misvaluation proxies indicates the extent to which a firm’s market value

deviates from its fundamental value. The difference between B and V is that, the former measures the

firm’s book value whilst the latter contains analysts’ forecasting information. Therefore, P/B is a

proxy for the firms’ growth opportunities or managerial effectiveness, which is used to investigate

the Q hypothesis, whilst P/V is more likely a pure proxy for misvaluation in this context given that

the measure reflects expectations of analysts’ forecasts about the firm’s future performance. However,


179

it is argued that analysts’ forecasts may not accurately reflect the information about the firm’s

prospects and managerial effectiveness. Therefore, the authors propose to include both P/B and P/V

to investigate the misvaluation hypothesis.

Dong et al.’s study (2006) focuses on the relationship between the valuations of both the bidder and

the target and many aspects of M&As, including the method of payment, M&A offer premiums, M&A

outcomes and a wide range of deal characteristics. The sample includes 2,922 successful and 810

unsuccessful takeover bids announced between 1978 and 2000, and filters out those M&A deals

whose transaction value was less than $10 millions, eliminating concerns about small deals

confounding the result.57 Dong et al.’s data show that bidders are generally those with a higher

valuation whereas targets are those with a lower valuation, reflected in higher P/B and P/V of bidders

opposed to targets. Once again, a higher valuation firm is more likely be the takeover bidder than a

lower valuation firm, which is consistent with the prediction in Shleifer and Vishny’s theoretical

model (2003). Stock-financed takeover bids show significantly higher valuation of the bidder and the

target than cash-financed takeover bids. Of particular note, the mean valuation differential between

the bidder and the target are highly significant in the two misvaluation measures in the stock-financed

acquisition subsample, whilst P/B has more explanatory power than P/V in explaining the difference

in bidder-target valuation in the cash-financed acquisition subsample. With respect to M&A outcomes,

the authors noted that bidders with higher P/B tend to pay targets with higher M&A offer premiums,

which are associated with negative market reactions to the bidder and positive market reactions to the

target.

57 The authors also winsorised P/B and P/V of the firms at 1% and 99% levels for a similar reason.


180

Their study proposes two different proxies for misvaluation and find some rather mixed results when

relating to a wide range of takeover characteristics. The authors found that cash bidders are also

relatively more overvalued than targets. The misvaluation effect triggers bidder managers’ incentives

of exploring mispricing targets. As a result, rather than paying targets with stocks, bidders who are

more overvalued would also pay cash to the undervalued targets. This argument complements stock-

market driven acquisitions, reasoning that either cash or stocks may be used when the bidders’ M&A

motive is to dilute overvaluation. The authors also note that these proxies for misvaluation are a

reflection of the managers’ perceptions of a firm’s valuation rather than the true position of the firms,

explaining relative valuations from a behavioural finance perspective. Despite that, the study has

thoroughly investigated the misvaluation hypothesis and provides empirical evidence relating to what

the theoretical model of misvaluation had proposed, failing to provide any rationale as to why target

managers accept overvalued stocks. It remains unclear whether target managers accept bidders’ stocks

believing that bidders with high P/B or P/V are those with good past performance and thus are able

to create synergies for the combined firm.

With this in mind, Rhodes-Kropf and Viswanathan (2004) and Rhodes-Kropf et al. (2005) proposed

that targets accept overvalued stocks is because they are misled by valuation errors. Whilst Burch et

al. (2012) suggest that targets prefer to hold highly valued stocks as they believe those firms are

normally well-run and have better prospects, implying that targets can identify the sign of

overvaluation and voluntarily accept those highly valued stocks. Their findings are consistent with

the prediction of the Q hypothesis. By accepting those highly overvalued stocks, targets could

liquidate their holding position.


181

Many additional studies have proposed various proxies for misvaluation, which challenge the

conventional misvaluation measures. For example, Ben-David et al. (2015), who used short interest

as a proxy for misvaluation, reflecting investors’ under- and over-reactions to the firm’s valuations.

The authors found that stock bidders have higher short interests relative to targets. The use of short

interest as a measure for misvaluation isolates the effect of growth opportunities which may lead to

the same prediction that stocks are likely to be used as a means of payment for M&As. Ben-David et

al.’s results are consistent with the misvaluation hypothesis. Di Guili (2013) challenges MTBV by

arguing that it represents both the misvaluation and growth opportunities of a firm. It is generally

believed that firms with a high MTBV may be a result of overvaluation or better investment

opportunities. So as to distinguish the effects of misvaluation and investment opportunities, the author

develops a new measure for investment opportunity, which is based on post-merger investment.

Analysing a sample of 1,187 mergers announced between 1990 and 2005 and using the average ratio

of capital expenditures over assets in the four years following the merger as a proxy for the firms’

investment opportunities, the author finds that the propensity for using stocks increases with the post-

merger investment ratio. His study indicates that the firm’s market value relative to its fundamental

value is a problematic proxy for misvaluation and likely to bias the predictions of the misvaluation

hypothesis.

4.2.2. Reference point theory of M&As

The reference point is an essential implication of Kahneman and Tversky’s prospect theory (1979).

It portrays the concept that people’ s feelings about the gain and the loss are judged relative to a

particular reference point, which is normally the current position of a subject, the status quo or the

aspiration level, as opposed to a reference point. Reference points distinguish prospect theory from


182

the traditional expected utility theory that emphasise the final outcome of the wealth level but fails to

account for the decision-making process.

The reference point theory is rooted in Tversky and Kahneman’s anchoring and adjustment

mechanism (1974) suggesting that people tend to rely heavily on a piece of salient but perhaps largely

irrelevant information at an initial stage of the decision-making process and adjust the final estimates

when new information is incorporated. This psychological phenomenon refers to the belief formation

process. Many implications of prospect theory are largely built upon the reference point theory. There

exists a loss-aversion tendency, given an equal-size of loss or gain, the feelings about the loss are

more pronounced than those about the gain (Kahneman and Tversky, 1979), and there also exists a

diminishing sensitivity that the feelings about the gain and loss are more pronounced in the scenario

where the outcome is close rather than distant in relation to the reference point (Tversky and

Kahneman, 1992).

The reference point theory interprets how people make decisions under the situation where

considerable information uncertainty exists. Baker et al. (2012) were the first to relate this theory to

the M&A pricing decision. They noted that the use of the target 52-week high as the reference price

lies in two proposed reasons. First, the firm’s 52-week high is frequently reported by the financial

media, which reinforces the investors’ formation process to utility realisation. Secondly, Baker et al.

justify the rationale behind the use of the target 52-week high from the psychological and practical

points of views of both the target and the bidder firms.58

58 The researcher discussed the main findings of their paper in the previous Chapters.


183

The authors also document evidence of the reference point effect on the market. When assessing the

deal outcome, the authors found that there are high offer premiums to targets who negotiate bidders

with their 52-week highs, whilst bidders receive negative market reactions, indicating that the offer

price driven by reference price is due to an overpayment. On the whole, Baker et al.’s study (2012)

suggests an evident reference point phenomenon for both the individual and the institutional investors.

Chira and Madura (2015) extended Baker et al.’s work (2012) by interpreting the role of reference

point prices as the bargaining power of the firm. Chira and Madura focused on the relationship

between takeover possibility and the reference points of the two firms involved. Similar Baker et al.’s

work (2012), Chira and Madura measured the reference point as the deviation of the firms’ 52-week

highest stock price from their current prices. The study assumes that the firms have greater bargaining

power when its current price is closer to, rather than distant from, its highest stock price over the year.

There results show that firms are likely to be takeover bidders when they have less distance between

their current price and the highest price over the year, indicating that a bidder’s great bargaining

power leads to a greater incentive of undertaking acquisitions. Moreover, it was found that firms are

less attractive to unaffiliated firms when the distance between the reference price and the current price

is large, arguing that the firm whose price deviates considerably from its 52-week high tends to resist

the deal unless a high price is offered in compensation. And because of this, bidders will abandon the

deal.

The authors suggest that both the target and the bidder will assess the firm value with their own

reference points. By examining the relationship between deal completion and the firms’ reference

points, the authors argue that targets with a low stock price relative to their 52-week highs may


184

perceive themselves as undervalued, resulting in strong deal resistance. Bidders will pursue the deal

by paying for takeover targets with stocks whose price is high relative to their 52-week highs, given

that bidders perceive their firms to be overvalued. However, their study fails to provide any rationale

on why bidders would look at their own reference points since they have more information about their

firm than other investors in the market. In addition, the study argues that the deal is likely to be

completed when the price distance between the firm’s 52-week high and the current price is small for

both the target and the bidder, relying on the explanation that psychological barriers between the two

firms are low.

Many additional studies have indicated an important role the firm’s reference point price plays on

investors’ trading strategies. George and Hwang (2004) used a firm’s 52-week high as the reference

point price and found evidence of reference dependence bias in the stock market. They note that

investors view the price that is close to its 52-week high price as good news and the price that has a

great distance its 52-week high as bad news. The mechanism is that investors are unwilling to bid up

a price when it is close to the 52-week high and should be reluctant to press down a price when it is

a long way below the 52-week high. This reasoning suggests that investors have a tendency to make

decisions based on a reference point price, which is the firm’s 52-week high. The salient evidence of

reference dependence is also supported by the findings of Grinblatt and Kelaharju (2001) who

investigated the Finnish stock market. They used a firm’s historical high as a reference point price

and found that investors avoid selling (buying) stocks whose value is far below (close) to the historical

high. Huddart et al. (2009) observed the phenomenon of abnormal sales’ volume around the 52-week

high based on a sample of 2,000 firms over 24 years.


185

Shefrin and Statman (1985) put forward the disposition effect that investors tend to sell winner stocks

too soon and hold loser stock too long, which is in favour of prospect theory. Barberies and Xiong

(2009) also documented evidence of loss-aversion based on a large sample of activities by individual

investors. These authors show investors’ greater propensity for selling stocks whose value has risen

recently and buying those whose value has been downgraded recently. According to a survey of

reference point theory (Olsen, 1997), it often occurs that the reference point effect is evident among

both individual and institutional investors, which implies that reference dependence bias is human

nature when facing uncertainty.

Bidder and target managers look at the reference point of the firms because they know the market

also has a reference-dependence bias. In this case, bidder managers can justify the motive for M&A

offer premiums paid for the target. Targets demand offer premiums around the target 52-week high

to avoid being blamed for selling the firm with a low valuation. Though Baker et al. (2012) showed

that offer premiums are positively related to the target 52-week high, it is less likely for the researcher

to know the whole reason for the bidders’ M&A motive without considerations about the bidder’s

side. For example, bidders are overconfident as they would argue that paying for targets based on

their 52-week highs can create synergies. It may also be the case that bidders are in a relatively weaker

bargaining position to targets because of information asymmetry, indicating that offer premiums

relative to the target 52-week high are due to target’s strong bargaining power. This chapter puts

forward Baker et al.’s idea (2012) by adding the role of the bidder reference point, and explores how

M&A activity is structured, based on the reference point of the two firms involved.


186

4.3. Hypothesis development

Shleifer and Vishny's misvaluation hypothesis (2003) indicates that bidders have a direct incentive to

use stocks as means of payment to finance an M&A deal when they are overvalued. It is the case that

overvaluation is a managerial perception as noted by Dong et al. (2006). The market has a tendency

to assess the firm value based on the reference point price, as predicted by reference point theory of

M&A (Baker et al., 2012). Managers whose aim is to time the market will also consider this and

eliminate such overvaluation concerns through M&As. Bidders are likely to be those with a stable

performance relative to targets, drawing its current price closer to the 52-week high, rather than the

case of the target. This gives rise three possible situations when the firm’s value has recently moved

toward to its 52-week high. If bidders are fairly estimated, nearness to the 52-week high price

indicates bidder’s strong bargaining power. The market believes that the firm is strong in making

profits for the combined firm, and as a result, bidders would pay fewer offer premiums. The second

scenario is that bidders are possibly undervalued even though there is nearness to the 52-week high

price given that the firm’s growth opportunities may be underestimated. The arrival of the information

at a later stage will push the firm’s price to a new 52-week high. If so, bidders would possibly pay

for targets with cash rather than undervalued stocks. The third scenario is that bidders are overvalued

with nearness to the 52-week high price. If it is the case, bidders are likely to pay for a target with

stocks, aiming at diluting overvaluation and creating firm value in the long run. Furthermore,

eliminating overvaluation requires additional costs, which result in a high offer price in excess of the

target’s current value.

The RRP eliminates any of these concerns on the bidders’ value, as it reflects the extent to which the

bidder is relatively more overvalued than the target from the market perspective. This proxy


187

investigates the valuation at the market level rather than at the firm level, which is the misperception

of the firm’s valuation created by those investors lacking private information about the firm. If

managers can time the market, M&As are a valuation game. It should be suggested that bidders should

dilute their valuations as long as they are relatively more overvalued than the target. One important

sign of this overvaluation is that bidders choose stocks as a means of payment for finance M&As,

since holding stocks in an overvalued market will destroy the wealth of long-term shareholders.

Therefore, the probability of using stocks for payment purposes increases in line with movements in

the RRP leading to our first testable hypothesis of

H1: There is a positive correlation between the RRP and the likelihood of using stocks as a means of

payment in M&As.

Shleifer and Vishny (2003) predicted that bidders’ overvaluation will be diluted through acquiring an

undervalued (or less overvalued) firm. Dong et al. (2006) proposed P/B and P/V as proxies for

misvaluation, which is the firm’s market value relative to its fundamental value. Dong et al.’s findings

imply that diluting overvaluation remains a priority objective for bidder managers regardless of the

method of payment used for finance M&As. It might be the case that cash bidders are also overvalued,

in that overvaluation drives acquisitions. According to this argument, bidders make a takeover bid

with cash rather than not bidding at all. Thus, bidders can profit from acquisitions by using stocks

when they are relatively more overvalued and using cash when targets are undervalued. This

reasoning implies that M&As result from bidder managers timing the market.

Relative bidder-target valuations were measured with RRP, capturing the market perception of a


188

firm’s valuation and also giving consideration of the valuation of the two firms involved. Since both

target and bidder managers will assess their firms’ value according to the reference point (Chira and

Madura, 2015), they can identify the signs of relative valuations. Thus, bidders would pay offer

premiums in exchange for the control of the target firm. Targets would demand high offer premiums

based on this relative valuation, as they can also identify any misvaluation reflected in the RRP. They

would resist the deal unless high offer premiums were offered in compensation. Based on the above

argument, it was tested that

H2: M&A offer premiums correlates positively with the RRP.

Finally, bidders with a relatively smaller changes related to the firm’s 52-week high (with a price that

is close to its 52-week high) are expected to have a stronger bargaining position, leading to low offer

premiums being paid to the target. Thus, a high offer premium would offer shareholders a

straightforward sense that there is an overpayment, thus reacting negatively to the bid announcement.

On the other hand, this researcher argues that M&As serve as a value enhancement opportunity for

the target firm as shareholders would expect that targets with a relatively higher reference point would

have a higher probability of profiting through acquisitions. Therefore, our last testable hypothesis is

H3: Bidder (Target) short-term performance correlates negatively (positively) with the RRP.


4.4.1. Data

The initial sample covers 36,506 U.S. domestic public acquisitions announced between January 1,


189

1985, and December 31, 2014, as provided by Thomson One. Stock price was collected from CRSP,

and a series of standard accounting variables were collected from COMPUSTAT. Accounting

variables were required to be available for the fiscal year end prior to the announcement date are

required. Public acquisitions refer to the two firms involved being publicly traded U.S. firms (listed

on NYSE/AMEX/NASDAQ).59 Once deals that were classified as recapitalizations, repurchases,

self-tender offers and rumors according to Thomson One were excluded, it was left with 11,615

observations.60 The offer premium is not a missing value, which further reduces the number of

observations to 5,450. The stock price for the calculation of the bidder and the target 3-day CARs

were required to be available, which left with 4,630 observations. The payment method information

to be available in Thomson One, which left the chapter with 4,290 observations. It yielded 2,156

observations after excluding all bidder variables with a missing value, and has a final sample of 1,878

observations after excluding all target variables with a missing value.61

The RRP effect on the probability of using stocks as a means of payment for acquisitions was analysed

by controlling for a series of deal, bidder and target characteristics, which are standard control

variables highlighted in M&A literature. Bidder size is expected to be negatively related to the stock-

financed acquisition. Faccio and Masulis (2005) suggest that larger bidders have higher credit

facilities, which reduces the probability for using stocks. The firm’s growth opportunities were

measured with MTBV. Higher MTBV bidders tend to use more stocks in acquisitions, in that they

59 Firms should have an available stock price to calculate the 52-week high.

60 By doing these, acquisitions in which the bidder and the target is the same firm were eliminated, on the rationale that

the difference between the target and the bidder reference points cannot be distinguished (e.g. self-tender offers), and

acquisitions in which the target actively searches for the bidder were also excluded, since those deals violate the reference

point effect. 61 All variables that have a missing value and were used in the regressions were excluded, following a summary of the

acquisition sample and variables.


190

reserve cash to fund new investment projects (Rhodes–Kropf et al., 2005; Dong et al., 2006). The

firm’s profitability was measured with the return-on-asset ratio (ROA). It should be suggested that

firms with higher profitability are more likely to use retained earnings held in cash rather than stocks

as it reduces costs of financing. Target characteristics were also accounted for, since stocks are more

likely to be used to mitigate the target risk (Hansen, 1987). In this respect, the propensity for using

stocks is greater when targets’ risks increase, such as large size, with a high MTBV and a low ROA.

Following Officer (2004), information asymmetry is measured by calculating the standard deviation

of returns. Hansen (1987) suggests that stocks are more likely to be used when level of information

uncertainty increases. Leverage is defined as debt-to-equity ratio (D/E). Vermaelen and Xu (2014)

suggest that over-levered bidders who justify stock financing in terms of moving to an optimal capital

structure lead to an increase of overvalued stocks for acquisitions. Whereas targets with high leverage

should be reluctant to receive such stocks. Liquidity is defined as cash flow-to-equity ratio (CF/E).

Higher liquidity firms are more likely to be less financially constrained firms, which result in the

method of payment for acquisitions is cash rather than stocks. Inclusions of capital structure related

variables in the regressions would allow us to disentangle the effects of firm’s capital structure

decision and misvaluation on stock-financed acquisitions.

In a further analysis, the RRP effect on the offer premium was analysed. Different categories of

variables were controlled in line with the work of Eckbo (2009). Specifically, firm size was measured

with logarithm of market valuation (MV). According to the hubris management hypothesis (Roll,

1986), larger bidders tend to pay generously for smaller targets. The firm’s profitability was measured

with ROA. Agency theory suggests that poor-performing bidders tend to dissipate firms’ resources


191

and overpay for the target, whereas well-performing firms are attractive to bidders (Schwert, 2000).

The firm’s growth opportunities were measured with MTBV. Rau and Vermaelen (1998) suggest the

extrapolation hypothesis that glamour bidders are less cautionary than value bidders about the target

valuation, leading to higher offer premiums. Harford (1999) suggests that the target MTBV links with

the managerial takeover motive since bidders are more aggressive in exploring synergies from the

lower MTBV target. Stock volatility were calculated with standard deviation of returns over 335

calendar days ending 30 calendar days prior to the announcement date. All regressions include year

and industry effect.

The M&A sample was matched to firms’ characteristics collected from COMPUSTAT, and to stock

price information collected from CRSP. By doing this, all relevant information was presented in one

data file for the purpose of conducting analyses. MV is defined as the product of market price and

outstanding shares (CRSP: SHROUT*PRC). “Relative Size” is defined as the deal value divided by

bidder MV. ROA is return-on-asset ratio, defined as net income (Compustat: NI) divided by total asset

(Compustat: AT). MTBV is market-to-book value, defined as the market value of equity to the book

value of equity, where book value of equity is total shareholders’ equity (Compustat: SEQ) plus

deferred taxes and investment tax credit (Compustat: TXDITC) minus the preferred stock redemption

value (Compustat: PSTKRV). Capital structure related variables were also included in the regressions

of the probability for using stocks on RRP. Leverage was measured by debt-to-equity ratio, defined

as total long-term debt (Compustat: DITT) divided by the book value of equity. CF/E is cash flow-

to-equity ratio, defined as income before extraordinary items (Compustat: IBC) plus depreciation and

amortization (Compustat: DPC) minus cash dividends (Compustat: DV).


192

4.4.2. Summary statistics

Table 4.1 reports a summary of the acquisition sample. Stock-financed acquisitions are those refers

to acquisitions that are 100% financed by stocks. Cash-financed acquisitions are those refers to

acquisitions that are 100% financed by cash. “Mix” refers to acquisitions that are neither all stock-

financed nor all cash-financed acquisitions. “Completed” refers to completed deals in the sample

period investigated in our chapter. “Tender” denotes tender offers, referring to deals that are proposed

directly to shareholders. This opposes to mergers, which made through negotiation of the

management teams of the two firms involved. “Hostile” denotes to hostile bids, portraying deal

attitude. “Diversification” refers to diversified deals in which the primary two Standard Industry

Classification codes are different between bidders and targets.

The average transaction value for 1,878 M&A deals in the sample is 1.54 billion US dollars. Of 1,878

acquisitions, 608 all stock-financed acquisitions, 726 all cash-financed acquisitions and 539 mixed

acquisitions.62 There were 702 diversifying acquisitions, as opposed to 1,176 relatedness deals, and

1,597 successful acquisitions compared with 277 unsuccessful deals.63 There were a small proportion

of tender offers and hostile acquisitions, 380 and 134 respectively.

Table 4.2 reports summary statistics for variables. Panel A presents the dependent variables used in

OLS regressions, including offer premiums, the bidder and the target 3-day CARs calculated using

the market model. The mean value for offer premiums is 31%. It is suggested that bidders have to pay

on average over 30% in excess of the target’s market price to gain the firm control. The mean value

62 The method of payment information for five acquisitions is defined as “Others” in Thomson One.

63 There were 1,874 deals with information about deal status based on Thomson One.


193

for bidder CARs is lower than target CARs, 22.2% over -1.1%. Panel B presents the main variables

of interests. The mean value for the bidder reference point is lower than the target reference point,

29.4% to 41.2%, which suggests that bidders on average are relatively more overvalued. The mean

value for RRP is 11.8%. Panel C presents all control variables. Bidders are on average larger than

targets, 7.50 as opposed to 5.21. Bidders have better investment opportunities than targets, 3.99

compared to 2.69. The profitability for bidders is higher than for targets, 2.9% over -4.9%, reflected

in ROA. Overall, firms with larger size, higher profit-generating ability and better investment are

likely the takeover bidders whereas those with smaller size, lower profitability, and lack of investment

opportunities are likely the takeover targets. The summary statistics are generally consistent with

prior M&A literature (Fuller et al., 2002; Moeller et al., 2004).

4.4.3. Methodology

4.4.3.1. Relative reference point (RRP) and offer premiums

Reference point (RP) refers to what extent the current price deviates from its 52-week highest price.

The 52-week highest price is a highly relevant piece of price information that shapes investors’ minds

to the firm’s prospects. Lacking information and time availability to process the deal information, the

market would naturally borrow this to compare with the current price. Hence, the firm’s 52-week

highest price is the market reflection of the firm’s best performance. According to George and Hwang

(2004), the 52-week highest price is the outcome of a series of good news that occurred in the past

that drives the firm’s market value beyond its fundamental value. Thus, the 52-week high is largely

related to the market’s perception of the firm’s valuation. A lower reference point would indicate that

the firm is still in the momentum of the “good news” effect, leading firms to be more overvalued. In

contrast, a higher reference point indicates that the good news effect is less relevant, leading firms to


194

be less overvalued. Therefore, a target reference point (TRP), larger relative to the bidder reference

point (BRP), sending a signal of market-wide perception regarding the target firm being less

overvalued than the bidder firm. Furthermore, the extent to which the bidder is more overvalued than

the target is measured with the relative reference point (RRP). Therefore, the market misperception

of firm’s valuation tends to reduce when a lower reference point firm mergers with a relatively higher

reference point firm. The data also show that TRP is on average larger than BRP.64 Formulas for RP,

RRP and offer premiums were illustrated as follows:

, 30 , 30log(52 ) log( )i i t i tRP WeekHigh StockPrice (4.1)

i i iRRP TRP BRP (4.2)

, , , 30log logi t i t i tOfferPremiums OfferPrice TStockPrice (4.3)

where iRP denotes the reference point of each firm i. The bidder (target) reference point is defined as

the logarithmic term difference between the bidder’s (target’s) highest stock price over 335 calendar

days ending 30 days prior to the announcement date and bidder’s (target’s) stock price 30 days prior

to the announcement date.65 iRRP denotes relative reference point which is defined as the target

reference point (iTRP ) and the bidder reference point (

iBRP ). Offer premiums were calculated as the

64 The difference between the target reference point and the bidder reference were used, attempting to obtain a positive

value, making the coefficients easier to interpret. According to the prediction, a less overvalued target should have a

higher reference point whereas a more overvalued bidder should have a lower reference point. The descriptive statistics

also show that, on average, the target reference point is larger than the bidder reference point: 0.412 and 0.294.

65 Using the firm’s stock price 30 days prior to the announcement date tends to mitigate the market reactions released due

to information leakage.


195

logarithmic term difference between the offer price ( ,i tOfferPrice ) and target stock price 30 days prior

to the announcement date ( , 30i tTStockPrice ).

As this chapter examines the effect of a relative valuation between the two firms involved on an

M&As transaction but not the extent to which the firm’s market value deviates from its true value,

the use of the RRP allows the researcher to establish a bilateral valuation framework where it is the

valuation deviation of the two firms influences the market perception. The market-wide valuation

would potentially drive the firm’s market value away from its fundamental value (Rhodes-Kropf and

Viswanathan, 2004). By doing so, it can account for the influence of market swings, as the RRP is a

relative valuation measure allowing the researcher to observe a new equilibrium of the firms’

valuation.

4.4.3.2. Logistic regressions

Logistic model refers to that dependent variable is categorical. The dependent variable of binominal

logistic model takes two values. This model was employed to study whether acquisitions are purely

financed by stocks or other means of finance, taking value of 1 and 0 respectively. While the

dependent variable of multinomial logistic model takes more than two values. The model was

employed to study the magnitude change in the RRP causes the likelihood of the acquisitions are

financed by stocks rather than cash or the mix payment. Both binomial and multinomial logistic

models were employed in this chapter. Binominal logistic regression was used to examine the RRP

effect on the probability of using stocks. The logistic coefficients were transferred into marginal effect

(ME) at sample means, which provide consistent interpretation of OLS estimates. In addition,

multinomial logistic regression was used to examine the RRP effect on the probability of using stocks


196

rather than other means of payment as a robustness check. The model was specified as follows:

, 1 , ,

1

( ) ( )N

i t i i i t i t

i

P Stock Pl RRP X

(4.4)

, , 1 , ,

1

( ) 1 ( ) ( )N

i t i t i i i t i t

i

Q Stock P Stock Pl RRP X

(4.5)

where Pl represents the possibility function, the main variable of interest to be investigated is iRRP .

iX denotes control variables in relation to the payment choices for M&A deals. (4.4) and (4.5)

combined yield the behavior of the logistic function, as follows:

'

1 , ,

1

( )N

i i i t i t

i

Pl RRP X

= ,( )i tP Stock ,( )i tQ Stock (4.6)

The function represents the derivative in respect of the ,i tRRP .

4.4.3.3. Classification of high-, neutral- and low-valuation markets

Rhodes-Kropf and Viswanathan (2004) suggest stock-financed acquisitions are positively correlated

with market-wide valuation. Following Bouwman et al.'s approach (2009), market valuation periods

were classified using the price-earnings (P/E) ratio of the market index (S&P 500) and monthly data.

Firstly, the market P/E ratio was de-trended by removing the best straight line fit (OLS) from the P/E

of the month in question and the five preceding years. Secondly, each calendar month was classified

into high- (low-) market valuation groups if the de-trended market P/E ratio of that month was above


197

(below) the five-year average. Then, the months were ranked according to the de-trended market P/E

ratio. Months in the top 25% of the above average group were classified as high-market valuation

months, months in the bottom 25% of the below average group were classified as low-market

valuation months, the remaining months being classified as neutral-market valuation months. Thus,

half of the months were classified as neutral-market valuation and the other half contains months of

both high- and low-market valuation. The idea of de-trending market valuation is to remove the

upwards trend because the most recent acquisitions generally have a higher market valuation than the

past due to market inflation.

4.4.3.4. Short-term method

Following Eckbo et al.'s market model (2016), firms’ announcement returns were calculated as

follows:

, 1 , ,i t m t i tR R (4.7)

where ,i tR denotes holding period returns (CRSP: RET) for firm i in the period t, ,m tR denotes value-

weighted market returns including dividends (CRSP: VWRETD), ,i t denotes the error term. We

estimate market model parameters over the window from 261 to 28 trading days prior to the

announcement date [-261, -28], and use a 3-day event window [-1, 1].

4.4.3.5. Long-term method

4.4.3.5.1. Market-adjusted model

Following Loughran and Vijh (1997), the firm’s long-term performance was calculated with the


198

market-adjusted buy-and-hold abnormal returns (BHARs). And 36-month BHARs were presented

with the following equation:

, , ,

1 1

(1 ) (1 )T T

i t i t index t

t t

BHAR R R

(4.8)

where, ,i tR is the arithmetic returns for firm i on day and ,index tR is the arithmetic return for the market

index on day t.

4.4.3.5.2. Size-adjusted model

Fama (1998) claims that the use of different long-term methodological approaches can generate

different results. He suggests the efficient market hypothesis that financial anomalies are chance

results due to the market misinterpretation of the information, such as under or over-reactions. They

are largely contributed to the bad-model problems. Fama (1998) suggests that the long-term return

anomalies tend to disappear with reasonable change in a way they measured. In the light of this, the

size-adjusted model was employed to assess bidders’ long-term abnormal returns. Lyon et al. (1999)

suggest that the use of this methodological approach can yield well-specific test statistics as it

alleviates the new listing and rebalancing biases. Following Lyon et al’s work (1999), BHARs were

measured by constructing reference portfolios based on all firms listed in NYSE. Specifically, size-

adjusted ARs were calculated as follows:

, ,

0

(1 ) 1T

i t i t pt

t

BHAR R R

(4.9)


199

where ptR relates to reference portfolio return and calculated as follows:

,

0

1

(1 ) 1T

j tnt

pt

j

R

Rn

(4.10)

where ,j tR relates to simple return on the firm j at time t, and n denotes to the number of firms.

4.4.3.5.3. Bootstrapped t-statistics

Fama (1998) challenges the statistics for the univariate test of BHARs under conventional

methodological approaches. He argues positive-skewness problems that may not yield well-specified

t-statistics. In the light of this, a bootstrapped t-statistic and the non-parametric Wilcoxon rank-sum

test were used to estimate the differentials for BHAR ranks. The skewness-adjusted t-test is presented

as follows:

21 1

3 6T N S S

N

(4.11)

where

1, 2( )

BHAR

BHAR T TS

(4.12)

and


200

1, 21, 2

3

( )( )

2

3

n

T Ti T T

i

BHAR

BHAR BHAR

N

(4.13)

where BHAR relates to the standard deviation of BHAR by conventional models.


Multivariate analysis is employed to examine the relevant factors explaining the market reaction.

According to Draper and Paudyal (2008), multivariate analysis is superior in analyzing the causation

relationship between the market reaction and related variables. The multivariate framework is

presented as follows:

( 1, 1)

1

N

i i i

i

CAR X

(4.14)

where iX denotes variables related to market reactions. The main variable of interest to be

investigated is RRP, which is the difference between the target and bidder reference points. The

multivariate framework controls for a series of deal and firm characteristics that have significant

impacts on market reactions in standard M&A literature (Alexandridis et al., 2010; Asquith et al.,

1983).


In this section, empirical results are reported and discussions being provided in relation to the RRP

effect on the method of payment, offer premiums and bidder performance in the short and long runs.


201

According to misvaluation hypothesis, overvalued bidders tend to time the market by paying

significantly high offer premiums to target shareholders and utilising their overvalued stocks. Despite

wealth being destroyed in the short run, the hypothesis predicts that rational bidders expecting to

create firm value in the long run. Following these predictions, logistic regressions of stock-financed

acquisition on the RRP were first conducted. Second, how offer premiums change according to RRP

is investigated. Our third main regressions examined the role of offer premiums plays on firms’

announcement returns. Specifically, the 2SLS technique is used to examine how announcement

returns of a firm change according to the RRP driven offer premiums. Finally, the post-merger period

performance of stock bidders is examined with univariate analysis by RRP ranking.

4.5.1. The RRP effect on the probability of using stocks

Table 4.3 reports a positive relationship between the RRP and the likelihood of using stocks for

financing M&As. The sign and significance level of the RRP do not change significantly after

information asymmetry has been included, as well as capital structure related variables of both the

bidder and the target.66 Specification (4) of Table 4.3 shows the main results, a one unit increase in

the RRP lead to about 10.4 percentage points increase of stock payments (p = 0.001), as interpreted

by marginal effect. In addition, an increase of MTBV results in an increase in the likelihood of the

use of stocks, suggesting that bidders tend to pay with stocks when they have better investment

opportunities on the reasoning that cash is likely to be retained for future investment. This is

consistent with Dong et al.’s view (2006) and that of Di Guili (2013). The results obtained show that

bidders who are small, with lower profitability and high debt are likely to pay with stocks, as they are

generally financially constrained which limits their cash borrowing. This is consistent with Eckbo et

66 Our results do not change significantly after considering the firm-fixed effect.


202

al.'s findings (2016).

The main results suggest that the relatively more overvalued bidders are likely to pay with stocks,

which is consistent with the misvaluation hypothesis and also reconciles the predictions of the

reference point theory of M&As. Bidders whose price is close to their 52-week high would give

targets a chance of selling out overvalued shares for profits following acquisitions, which is consistent

with Burch et al.’s view (2012) that targets tend to accept overvalued stocks because that bidders can

generate profits in the future. Alternatively, targets would demand more stocks from bidders when

targets can identify overvaluation, which is in line with Vijh and Yang’s view (2014). When the

market news has driven the target firm’s current value away from its fundamentals, bidders have the

incentive to exploit the target price discounts. With the reference-dependence bias, bidders also assess

the M&A valuation by focusing on the RRP, arguing that if a target’s price falls greatly below its 52-

week high than that of their firms, would offer bidders greater potential for overvaluation dilution.67

Table 4.4 reports the RRP effect on the probability of using stocks under different market conditions.

Rhodes-Kropf and Viswanathan (2004) suggest that stock-financed acquisitions are positively

correlated with high market-wide valuation. This is because that the overestimation of synergies

increases with valuation errors. Consistent with this view, it is found that the RRP effect is more

pronounced when market valuation errors are sizable. As explained by the marginal effect in Table

4.4, for every one unit increase in the RRP would lead to an increase of about 15.3 percentage points

67 It can be argued that when the target current price is significantly lower than its 52-week high, the target may experience

risks of bankruptcy. It is believed that bidders would be cautious about this and may not focus on the target 52-week high.

However, these individual cases were not considered in this thesis, as there is only limited number of deals in this M&A

sample.


203

when the market-wide valuation is high and an increase of about 10.9 percentage points when the

market-wide valuation is neutral, whereas the RRP effect on market condition is insignificant in the

subsample of low market-wide valuation.

4.5.2. The RRP effect on the offer premium

The view that investors’ perception is captured by the reference point effect has been examined. It

remains interesting to explore the RRP effect on offer premiums. Baker et al. (2012) show that the

market would look at the firm’s reference point and may perceive that an offer price is a result of

bidder’s overconfidence, given limited information and limited time in which to process the deal. It

is less likely that the reference point effect plays a similar role for the bidders who have more

information about the firm and seems less likely to justify a firm’s valuation according to a single

number.

The first four specifications of Table 4.5 use different categories of control variables.68 The sign and

significance level of the RRP do not change significantly compared with what is reported in the

specification (5), which is the main specification. The RRP is positive and significant at the 1% level

(coefficient 0.089, t = 4.501), showing that a 10% increase in the RRP is associated with an

approximately 0.9% increase in offer premiums. The signs and significance levels of the control

variables are generally consistent with prior M&A studies (Moeller et al., 2004; Alexandridis et al.,

2013). The results obtained by the author suggest that the relatively more overvalued bidders overpay

68 Specification (1) reports the relationship between offer premiums and the reference point, specification (2) controlled

for deal characteristics, specification (3) controlled for deal and bidder characteristics, specification (4) controlled for deal

and target characteristics, specification (5) controlled for all variables. All specifications accounted for year and industry

dummies.


204

for the target according to the RRP, which is consistent with the reference point theory of M&A. The

joint bidder and target reference point effect has an impact on target pricing. It can be interpreted as

follows: bidders who risk being perceived as overvalued have to pay for targets according to the target

reference point to obtain the deal before revising market perceptions, since lower than the target

reference point would decrease the probability of deal success (Baker et al., 2012). Moreover, when

the bidder’s price is close to the 52-week high, managers may find it hard to justify, and will suffer

significant losses in the long run either because of overvaluation to be corrected in the market (Jensen,

2005) or their stocks are to be aggressively sold at around the high market value of the firm (Barberis

and Xiong, 2009). Hence, bidders would pay heavily to revise the market’s misperception. On the

other hand, targets also demand offer premiums based on the bidder reference point, as a lower bidder

reference point might lead targets to believe that bidders are able to afford a higher offer premium.

Given that acquisitions were largely driven by operating synergies and agency problems before 1990,

and by misvaluation after 1990 (Shleifer and Vishny, 2003; Dong et al., 2006), time-distribution tests

whose results are reported in Table 4.6 were carried out. The sample period in this chapter divides

into three based on M&A merger waves. Before 1990, 1990 to 2000 and post 2000. The sample period

covers the high valuation trends when investors’ perception to firms’ misvaluation is high, such as

the stock market bubble between 2000 and 2002, and the housing bubble and credit crisis between

2007 and 2009. It should also be suggested that the misvaluation hypothesis is more likely to explain

acquisitions after 1990 than before 1990 when the primary M&A motive is synergies. The results

yield similar results to our predictions. By extending the sample period of Dong et al. (2006), it was

found that RRP plays an important role during large valuation trends. The sample period divides into

two: from 2001 to 2007 and 2008 to 2014, in order to investigate how RRP explains the offer


205

premiums before and after 2008 financial crisis, as reported in Table 4.7. The results obtained indicate

that for every 10% increase in the RRP an increase of 0.94% in the period of 2001 and 2007 would

occur, this figure more than double during the period of 2008 to 2014. Misvaluation tends to be larger

after financial crisis. Once again, the results the research obtained show that the relative overvaluation

drives M&A overpayment. Unlike Rhodes-Kropf and Viswanathan (2004), who found a positive

correlation between stock-financed acquisitions and merger waves and argue that rational targets

make mistakes by accepting overvalued stocks during high valuation periods, up-to-date evidence is

provided here suggesting that premiums increase with valuation errors. This also rationalises the

bidder’s motive for overpayment, as high offer premiums area form of compensation for the target’s

willingness to accept overvalued stocks.

However, it is important to be aware of the fact that not all acquisitions involve a bidder that is

relatively more overvalued than the target, as reflected in the RRP. The proxy is used to investigate

whether the primary M&A motive is overvaluation from the market perspective. Bidders are

relatively more overvalued if their current price shows smaller changes relative to the reference point

price as opposed to those of the targets (i.e. RRP>0), and they are relatively undervalued if the

changes are larger relative to the reference point price compared with those changes of targets (i.e.

RRP<0). If the RRP is a suitable proxy for misvaluation, it should accommodate the predictions of

the misvaluation hypothesis. Based on this, higher RRP acquisitions associated with higher offer

premiums and target announcement returns, and lower bidder announcement returns than lower RRP

acquisitions, should be expected.

The sample was divided into two subsamples according to the RRP. First, the univariate analysis


206

Results of Table 4.8 are presented. The results show a majority of deals are RRP driven: 1,155 as

opposed to 723. Panel A indicates that acquisitions are carried with significantly larger offer

premiums when bidders are relatively more overvalued than the opposite case, the mean difference

for the offer premium is 5% and at the 1% significance level. Our results suggest overvalued bidders

are likely to pay with high offer premiums for the undervalued or less overvalued targets. Managerial

priority in takeovers is to dilute overvaluation. Because of this, they believe the deal may not be

accepted until high offer premiums paid to targets who can also identify bidders’ motive through RRP.

In Panel B, both the two subsamples show significantly negative bidder announcement returns. The

relatively more overvalued bidders perform significantly worse than the relatively more undervalued

bidders. With respect to target announcement returns, both relatively more undervalued and more

overvalued targets receive high market reactions, which is significantly different from 0. In particular,

the mean difference for target announcement returns is about 5%, suggesting that targets involved in

high RRP acquisitions can demand high offer premiums based on RRP, which is translated into higher

announcement returns.

Panel D of this table reports univariate analysis results for the offer premium by the method of

payment in the two RRP subgroups. The results obtained are consistent with the prediction of the

misvaluation hypothesis that overvalued stocks are used as cheap currency. Specifically, all cash-

financed acquisitions carry higher offer premiums when bidders are relatively more overvalued than

relatively more undervalued bidders, as the result shows a mean difference of 8.5%, which is at 1%

significance level. Moreover, all stock-financed acquisitions carry lower offer premiums than all

cash-financed acquisitions when bidders are relatively more overvalued. Combined, relatively more

overvalued bidders tend to avoid using cash, as it will increase bidders’ takeover costs. Our results


207

indicate that the method of payment indicates clear sign that managers time the market.

Following univariate analyses, multivariate analyses of the offer premium on RRP in two different

RRP subsamples were conducted, as reported in Table 4.9. Specification (1) shows the offer premium

increases 1.64% for every 10% increase of RRP when the bidders are relatively more overvalued.

The results are not significant in the comparable analysis. Our results show bidders to be perceived

as relatively more overvalued are more likely to pay for targets with high offer premiums for the

purpose of revising the market misperception. Our results also show that larger bidders generously

pay for smaller targets, which is consistent with the findings of Moller et al. (2004).

4.5.3. Do bidders who focus on the RRP overpay for the target?

The RRP was used as an instrument variable of the offer premium when examining the role of

overpayment. Baker et al. (2012) suggest the offer premium is not a clean measure since it can

represent both overpayment and synergies. 2SLS enables the investigation of the chain responses of

the RRP effect on the offer premium, and the offer premium effect on firm’s announcement returns.

If the RRP plays a role in overpayment, it should be expected that the offer premium using 2SLS

estimates should yield more negative (positive) bidder (target) announcement returns compared with

those using OLS estimates.

In Table 4.10, the offer premium with 2SLS estimates generates more negative (positive) bidder

(target) CARs compared with that with OLS estimates, shown in specifications (1) and (2), and (3)


208

and (4) respectively.69 The results are consistent with Baker et al. (2012) who studied the reference

point effect on M&A outcomes in a sample of M&As between 1984 and 2007. The results are also in

line with the reference point theory of M&As. The market would presumably believe that the chance

of price rebounds tends to increase when the bidder’s current stock price is close to the 52-week high.

During the time when a bid is announced, the market reacts negatively to it as this may indicate that

bidders are unable to deliver real support to the firm’s performance and they are likely to undertake

bad acquisitions to maintain the overvaluation (Jensen, 2005).

4.5.4. Do all stock-financed acquisitions driven by the RRP protect the wealth of long-term

shareholders?

Thus far, the proposition that relatively more overvalued bidders are more likely to use stocks as a

means of payment and tend to pay significantly higher offer premiums using the new misvaluation

measure of the RRP has been examined. Now the issue of whether or not bidders focusing on RRP

protect long-term shareholders’ interest will be investigated. According to the misvaluation

hypothesis, bidders who dilute overvaluation with stocks attempt to protect the wealth of long-term

shareholders. Thus, it should be reasonable to expect that bidders making an offer price based on the

RRP. In Panel A of Table 4.11, the sample was limited to acquisitions that are 100% financed by

stocks only and the sample was ranked into four quartiles according to the RRP, each presenting 152

observations.70 The aim was to examine whether overpayment leads to underperformance. By doing

so, the offer premiums and long-term performance under the market-adjusted model for each

correspondent quartile were estimated. The fourth quartile (i.e. the highest quartile) includes

69 The Hausman test and F-test results show that coefficients generated by 2SLS regression are more consistent with those

generated by OLS regression, as reported at the end of Table 4.10.

70 Of the 608 all stock-financed acquisitions, 402 fall into the group in which bidders are relatively more overvalued.


209

acquisitions involving relatively more overvalued bidders whereas the other quartiles include

acquisitions involving relatively less overvalued or more undervalued bidders.

It was also found that stock-financed acquisitions generate negative long-term returns, which are

consistent with the M&A literature (Loughran and Vijh, 1997; Rau and Vermaelen, 1998). Bidders in

the highest quartile pay the highest offer premiums compared with those of other quartiles. In spite

of this, the mean difference of offer premiums between the relative more overvalued bidders and

undervalued bidders is 9.9% at 1% significance level and the mean difference for long-term abnormal

returns of stock bidders is 18.7% at 10% significance level, indicating that relatively more overvalued

bidders outperform their undervalued counterparts. Combined, this researcher’s results suggest that

bidders paying high offer premiums according to the RRP are able to protect the wealth of long-term

shareholders. These results are consistent with Ang and Cheng (2006) who found that the long-term

performance of stock bidders and overvaluation are positively related. However, the researcher’s

findings contradict to those of Lin et al. (2011). Their paper classified bidder valuation by the ratio

of price-to-fundamental value (P/V) with higher P/V indicating a more overvalued bidder, and found

that bidders who have the highest P/V generate a significant negative market performance for short

term and long term within three years following M&As as compared with those bidders in the other

P/V quartiles.

For the purpose of robustness checks, the long-term performance of stock bidders with the size-

adjusted model was also examined, and stock bidders were divided into three ranks according to the

RRP, as shown in Panel B of the table. This researcher’s results show that the relatively more

overvalued bidders tend to pay 36% offer premiums for targets, which is significantly 6.6% higher


210

than those paid by relatively more undervalued bidders. It is also evident that the long-term

performance of those paying significantly higher offer premiums is better. The results are quite similar

to those presented in Panel A of the table.


4.6.1. Endogeneity issues

OLS can be subject to endogeneity issues arising from omitted variable biases in this chapter, as the

RRP maybe correlated with firms’ mismanagement or mispricing which cannot be observed or the

possibility that the market perception is likely the be an accurate reflection of the firm’s valuation. If

the market could accurately estimate the value of the firm, the managers should have no chance to

time the market by means of the RRP. However, this tends to be unrealistic as the misvaluation

hypothesis proposes. In this case, in this paper it is suggested that the market’s 52-week high and the

bidder’s and target’s 65-week high can be used as instrumental variables given that they are indirectly

related to offer premiums but can affect offer premiums through the RRP. The market’s 52-week high

is an ideal instrumental variable as it is uncorrelated with a firm’s mismanagement. It is suggested

that the bidder’s and the target’s 65-week highs are used as instrumental variables in line with the

extrapolation hypothesis that indicates that market perception is influenced by firms’ past

performance. Using a longer horizon is of less relevance in terms of the market perception of a firm’s

valuation. A 2SLS estimator was therefore performed, with the market 52-week high and the bidder

and the target 65-week high as instrumental variables.


211

According to the results obtained, shown in Table 4.12, the OLS is preferred over the 2SLS.71 This is

because the market 52-week high reflects the market-wide valuation instead of the firm-specific

valuation which is believed to be an important source of valuation error (Rhodes-Kropf et al., 2005).

The results indicate that the OLS is likely to dominate the 2SLS. In addition to this, a correlation

matrix was performed, as reported in Table 4.13. The obtained results show little evidence of

econometric problems, such as multicollinearity issues.

4.6.2. The effect of the RRP on the probability of using stocks

Results of a series of robustness checks examining the RRP effect on the probability of using stocks

are reported in Table 4.14. Shleifer and Vishny (2003) predicted that bidders use stocks when they

are relatively more overvalued. The M&A sample in this chapter was divided into two sub-groups:

the relatively more overvalued bidders (i.e. RRP>0), and the relatively more undervalued bidders (i.e.

RRP<0).72 The results obtained suggest that the RRP effect on the probability of using stocks is solely

driven by the relatively more overvalued bidders. It was found that, when the bidders are relatively

more overvalued, every one unit increases in the RRP increases the propensity for using stocks for

acquisitions by 16.4 percentage points, whereas there is little evidence that the relatively more

undervalued bidders use stocks for acquisitions. Table 4.15 reports that the probability of using stocks

rather than other means of payment is also large when the RRP increases. As seen, stocks are more

likely to be used as a means of payment for acquisitions compared with the method of payment as

71 Specifically, the Hausman test shows a p-value of 0.6414, indicating there are no endogeneity issues in the regression.

Moreover, the test of over-identifying restrictions further rejects the null hypothesis that there is no relation between the

instruments and the error term (p = 0.0000).

72 1,742 observations in regressions were recorded. Of these observations, 1,103 fall into the group in which bidders are

relatively more overvalued and 639 fall into the group in which bidders are relatively more undervalued. 50 observations

were dropped due to multicollinearity problems in the industries.


212

cash and a mixture of stocks and cash respectively. 73 The results of additional controls are consistent

with the main results of this chapter presented in Table 4.3.

4.6.3. The effect of the RRP on the offer premium

This section reports the results of subsample robustness tests relating to the RRP effect on the offer

premium. The sample was divided into subsamples according to the method of payment, deal type,

deal choice and deal status in Table 4.16 to 4.19 respectively, and control variables are the same as

those in Table 4.5. Table 4.16 reports a positive relationship between the RRP and the offer premiums

by different methods of payment subsamples. The offer premium is larger when paying with cash

than stocks, which is consistent with the univariate analysis results presented in Panel D of Table 4.8.

It is worth noting that in stock-financed subsample, bidders tend to pay generously for those with a

higher ROA, implying that bidders tend to select firms with strong profit-generating ability, likely to

create value for the combined firm. The results obtained show that bidders are likely to pay with large

offer premiums regardless of the method of payment choices. Table 4.17 reports that an increase in

the distance of the RRP increases the offer premium regardless of deal types. Signs and significance

levels for additional controls are quite similar between the two subsamples. In Table 4.18, the offer

premium for merger subsamples is larger than for tender offers. It can be argued that a negotiation

process between the two firms involved may incorporate more private information about the firms,

leading the RRP to be well interpreted. In that case, bidders can more easily justify their M&A motives

as well as the target value with target managers rather than shareholders, thus resulting in lower offer

premiums. The sample used also includes a small proportion of unsuccessful deals as opposed to

73 Our results continue to hold by replacing the binary variable with a continuous variable. This allows us to observe how

the RRP affects the proportion of stocks used in the payment structure. We omitted this table for the sake of brevity.


213

successful deals. In Table 4.19, the RRP effect is strong for both subsamples.74 On the whole, the

results indicate that the RRP have a strong effect on offer premiums.

4.6.4. Do bidders who focus on RRP overpay for the target? Testing with the market-adjusted model

So as to avoid measurement error due to the model employed for the firms’ announcement returns in

analysing the role of the RRP played M&As, the market-adjusted model was used to calculate the

firms’ announcement returns, as reported in Table 4.20. These results are consistent with those in the

market model. Specification (1) reported that bidders’ CARs decrease by 1.4% for a one percent

increase in the offer premium, which is slightly higher than those predicted by the market model (-

1.8%). On the other hand, specification (3) indicated that targets’ CARs increase by 46.9% for every

one percent increase in the offer premium, which is as the same as that predicted by the market model.

A 2SLS estimator was also conducted using the RRP as an instrumental variable of the offer premium

in the regression of CARs on the offer premium, with a similar reason has been explained in the main

regression analysis. Including the RRP effect in the 2SLS, the results show a more positive bidder

CARs and more negative target CARs as compared with those under OLS. Once again, paying

according to the RRP impresses the investors’ minds that the deal is overpaid.

4.7. Conclusion

This chapter has investigated the misvaluation hypothesis using the reference point theory of M&As.

A bilateral valuation framework with RRP has been built, and results are consistent with what the

misvaluation hypothesis of Shleifer and Vishny (2003) predicts.

74 This was used as a robustness test rather than to interpret the RRP effect as the reason for deal completion, as what

determines a deal completion is complex.


214

The target and the bidder reference points have been added into the M&A platform and have found

that the propensity for paying for acquisitions with stocks is greater when the RRP increases, this

trend being more pronounced when the market misperception of a firm’s valuation is high. The results

obtained are consistent with those of Rhodes-Kropf et al. (2005). Moreover, the offer premium

increases with RRP, leading to more negative bidder (positive target) announcement returns. The

results indicate that RRP plays an overpayment role. In a quartile analysis for a sample of all stock-

financed acquisitions, it can be found that the relatively more overvalued bidders who pay highest

offer premiums compared with those in other quartiles generate less negative long-term abnormal

returns, suggesting that bidders time the market while focusing on the RRP.

Several contributions have been made to behavioural finance and M&A literature. First, a simply way

of structuring M&A through the RRP was built, which reflects the most current market reactions to a

firm’s valuations. This valuation measure is straightforward in terms of observing the difference

between bidder and target overvaluation. The market tends to react to bidders’ announcements

negatively, as there is a high offer premium paid according to the RRP. Secondly, it has been found

that the RRP directs the method of payment choices, as a relatively more overvalued bidder tends to

reduce offer premiums when financing an acquisition with stocks rather than with cash. Thirdly, it

has also been revealed that bidders who have more information about the firm are also subject to the

reference point effect. However, this does not mean that bidders are irrational, rather, the findings

show that higher offer premiums according to the RRP would mitigate negative market reactions in

the long term, suggesting that focusing on the RRP is a bidder’s way of thinking weighed towards the

interests of the long-term shareholders. Overall, the results of this chapter suggest that managers are

able to time the market through the RRP, which is consistent with the misvaluation hypothesis.

Chapter 4. Tables

215


This table reports summary statistics for 1,878 U.S. domestic public deals announced between 1985 and 2014. The number

N denotes the number of deals per year. The third and fourth columns present the mean and median of deal value. The

fifth to seventh columns present the method of payment. Here “Stock” (“Cash”) refers to all-stock (all-cash) acquisitions.

“Mix” refers to acquisitions that are neither all stock-financed nor all cash-financed acquisitions. “Completed” refers to

completed deals (i.e. successful deals), of which there are 1,874 with information relating to deal status. “Tender” refers

to tender offers. “Hostile” refers to hostile bids. “Diversified” refers to diversified deals in which the primary two Standard

Industry Classification codes are different between bidders and targets.

Year N Deal Value ($mils) Payment Method Completed Tender Hostile Diversified

Mean Median Cash Stock Mix Y N Y N Y N Y N

1985 5 243.86 53.00 3 1 1 3 2 - 5 1 4 2 3

1986 21 129.04 41.70 18 1 2 18 3 7 14 1 20 15 6

1987 34 254.60 47.25 18 6 8 30 3 6 28 5 29 17 17

1988 30 381.53 68.92 15 8 7 24 6 11 19 6 24 16 14

1989 24 114.04 30.49 11 10 3 17 6 4 20 1 23 18 6

1990 23 579.03 29.38 12 7 4 18 5 5 18 3 20 10 13

1991 23 172.38 26.82 7 14 1 19 3 3 20 - 23 14 9

1992 21 155.46 51.44 6 13 2 17 4 1 20 2 19 12 9

1993 45 519.07 114.00 14 18 12 33 12 7 38 5 40 20 25

1994 61 222.95 74.12 23 30 8 46 14 9 52 7 54 18 43

1995 98 538.98 74.91 27 53 18 84 14 14 84 6 92 38 60

1996 95 684.53 138.25 31 40 24 80 15 18 77 9 86 36 59

1997 130 645.10 232.11 19 62 49 113 17 19 111 3 127 46 84

1998 138 1208.89 140.12 36 54 47 126 12 27 111 3 135 47 91

1999 164 1513.65 305.42 61 58 45 134 30 41 123 14 150 68 96

2000 136 2286.88 378.34 35 68 33 119 17 32 104 8 128 49 87

2001 109 1115.01 146.89 33 40 36 94 15 27 82 4 105 38 71

2002 49 1784.72 268.90 20 14 15 44 5 16 33 4 45 18 31

2003 71 807.01 130.82 27 18 26 65 6 19 52 5 66 19 52

2004 66 2859.54 479.02 25 16 25 60 6 6 60 3 63 22 44

2005 66 2874.25 500.75 29 13 24 60 6 7 59 5 61 24 42

2006 75 1838.00 563.07 40 12 23 65 10 6 69 4 71 29 46

2007 60 1478.29 792.51 39 6 15 53 7 12 48 2 58 18 42

2008 57 2208.95 234.26 35 6 16 40 17 14 43 10 47 17 40

2009 44 3498.35 496.88 17 8 19 41 3 16 28 - 44 18 26

2010 56 1884.97 572.72 33 9 14 47 9 16 40 5 51 15 41

2011 41 2691.37 611.62 16 8 17 28 13 8 33 9 32 13 28

2012 38 1385.71 622.51 26 1 11 37 1 10 28 - 38 17 21

2013 44 1997.05 1139.09 28 5 11 39 5 9 35 3 41 12 32

2014 54 6908.66 1662.39 22 9 23 43 11 10 44 6 48 16 38

Total 1878 1540.29 227.49 726 608 539 1597 277 380 1498 134 1744 702 1176

Chapter 4. Tables

216


This table reports the number, mean, median and standard deviation of variables used in the regressions. Panel A presents the main

dependent variables. Firms’ 3-day CARs are calculated with the market model, with parameters estimated between 261 and 28 trading

days prior to the announcement date. Offer premiums are defined as the logarithmic term difference between offer price and target

stock price 30 calendar days prior to the announcement date. Panel B presents the main variables of interest. The reference point is

defined as the logarithmic term difference between a firm’s highest stock price over 335 calendar days ending 30 days prior to the

announcement date and the price ending 30 days prior to the announcement date. RRP is defined as the difference between the target

and the bidder reference point. Panel C presents control variables. Deal characteristics are noted as Table 1. “Relative Size” is defined

as the deal value divided by bidder MV, where bidder MV is defined as the product of market price and outstanding shares (CRSP:

SHROUT*PRC). ROA is return-on-asset ratio, defined as net income (Compustat: NI) divided by total asset (Compustat:AT). MTBV

is market-to-book value, defined as the market value of equity to the book value of equity, where book value of equity is total

shareholders’ equity (Compustat: SEQ) plus deferred taxes and investment tax credit (Compustat: TXDITC) minus the preferred stock

redemption value (Compustat: PSTKRV). Volatility is the standard deviation of daily returns over the 335 calendar days ending 30

days prior to the announcement date. CF/E is cash flow-to-equity ratio, defined as income before extraordinary items (Compustat: IBC)

plus depreciation and amortization (Compustat: DPC) minus cash dividends (Compustat: DV), and leverage is measured by debt-to-

equity ratio, defined as total long-term debt (Compustat: DITT) divided by the book value of equity. All accounting variables were in

the fiscal year end before the announcement date, and continuous variables are winsorised at the 1% and 99% level.

Variables N Mean Median Std. Dev.

Panel A: Main Dependent Variables

Offer Premiums 1878 0.310 0.292 0.282

Bidder 3-day CARs 1878 -0.011 -0.007 0.073

Target 3-day CARs 1878 0.222 0.176 0.240

Panel B: Main Variables of Interest

Bidder Reference Point 1878 0.294 0.157 0.351

Target Reference Point 1878 0.412 0.255 0.455

RRP 1878 0.118 0.055 0.420

Panel C: Other Variables: deal, bidder, and target characteristics

Cash 1878 0.387 - 0.487

Stock 1878 0.324 - 0.468

Hostile 1878 0.071 - 0.257

Tender 1878 0.202 - 0.402

Diversification 1878 0.374 - 0.484

Relative Size 1878 0.401 0.191 0.555

Bidder lnMV 1878 7.500 7.443 2.159

Bidder MTBV 1878 3.991 2.632 6.269

Bidder ROA 1878 0.026 0.047 0.138

Bidder Volatility 1878 0.029 0.025 0.017

Bidder Leverage 1872 0.661 0.330 1.194

Bidder CF/E 1804 0.161 0.180 0.316

Target lnMV 1878 5.212 5.133 1.839

Target MTBV 1878 2.691 1.797 4.300

Target ROA 1878 -0.049 0.024 0.241

Target Volatility 1878 0.041 0.035 0.022

Target Leverage 1870 0.619 0.140 1.621

Chapter 4. Tables

217

Table 4.3: The effect of the RRP on the probability of using stocks

This table reports binomial logistic regression for all-stock acquisitions on RRP. Dependent variable is “Stock”, which is a dummy variable, taking a value of 1 if acquisitions are 100%

financed with stocks, 0 otherwise. Variable definitions are as in the notes to Table 4.2. P-value is reported in parentheses. Statistical significance at the 1%, 5% and 10% levels is

denoted ***, ** and * respectively, is reported alongside marginal effects. We transfer coefficients into marginal effect (ME), evaluated at the sample means of the independent variables.

(1) (2) (3) (4)

Stock Coef. p-value ME Coef. p-value ME Coef. p-value ME Coef. p-value ME

RRP 0.567*** (0.000) 0.103 0.605*** (0.000) 0.110 0.531*** (0.001) 0.097 0.570*** (0.001) 0.104

Hostile -0.777** (0.014) -0.141 -0.742** (0.019) -0.135 -0.721** (0.022) -0.132 -0.670** (0.033) -0.123

Tender -2.441*** (0.000) -0.443 -2.402*** (0.000) -0.436 -2.563*** (0.000) -0.468 -2.502*** (0.000) -0.458

Diversification 0.146 (0.266) 0.027 0.124 (0.349) 0.023 0.146 (0.280) 0.027 0.120 (0.378) 0.022

RelativeSize -0.611*** (0.000) -0.111 -0.662*** (0.000) -0.120 -0.561*** (0.000) -0.102 -0.606*** (0.000) -0.111

Bidder lnMV -0.379*** (0.000) -0.069 -0.290*** (0.000) -0.053 -0.402*** (0.000) -0.073 -0.314*** (0.000) -0.057

Bidder MTBV 0.048*** (0.000) 0.009 0.036*** (0.001) 0.007 0.060*** (0.000) 0.011 0.048*** (0.000) 0.009

Bidder ROA -2.063*** (0.000) -0.375 -0.914* (0.088) -0.166 -2.464*** (0.000) -0.450 -1.407** (0.030) -0.257

Target lnMV 0.315*** (0.000) 0.057 0.350*** (0.000) 0.063 0.348*** (0.000) 0.064 0.375*** (0.000) 0.069

Target MTBV 0.051*** (0.001) 0.009 0.041*** (0.006) 0.008 0.063*** (0.000) 0.012 0.055*** (0.001) 0.010

Target ROA 0.087 (0.762) 0.016 0.371 (0.231) 0.067 0.009 (0.977) 0.002 0.218 (0.492) 0.040

Target Volatility 5.118 (0.253) 0.928 3.773 (0.423) 0.690

Bidder Volatility 32.029*** (0.000) 5.809 30.994*** (0.000) 5.670

Bidder Leverage -0.232*** (0.001) -0.042 -0.211*** (0.002) -0.039

Target Leverage -0.126*** (0.003) -0.023 -0.121*** (0.005) -0.022

Bidder CF/E 0.213 (0.403) 0.039 0.233 (0.386) 0.043

Year Yes Yes Yes Yes

Industry Yes Yes Yes Yes

Constant 0.324 (0.577) -1.480** (0.032) 0.592 (0.360) -1.130 (0.132)

N 1878 1878 1792 1792

Pseudo R2 0.272 0.287 0.291 0.303

Chapter 4. Tables

218

Table 4.4: The effect of the RRP on the probability of using stocks under different market-wide valuations

This table reports binomial logistic regression for all-stock acquisitions on RRP by different market conditions. Dependent variable is “Stock”, which is a binary variable, taking value

of 1 if acquisitions are financed with 100% stocks, 0 otherwise. High Market is a dummy variable, taking a value of 1 if takeover months in the top 25% above past 5-year average de-

trended P/E of the market index (S&P 500) or market valuation is high, 0 otherwise. Specifications (1)-(3) report the results when market-wide valuation is high, low and neutral

respectively. Due to the availability of the P/E of the market index, we were able to determine 1733 observations with valid market-wide valuation data, 666 for high valuation periods,

434 for low valuation periods, and 633 for neutral valuation periods. Variable definitions are as in the notes to Table 4.2. P-value is reported in parentheses. Statistical significance at

the 1%, 5% and 10% levels is denoted ***, ** and * respectively, is reported alongside marginal effects. We transfer coefficients into marginal effect (ME), evaluated at the sample

means of the independent variables.

Chapter 4. Tables

219

Table 4.4. (continued)

(1) (2) (3)

Stock High Low Neutral

Coef. p-value ME Coef. p-value ME Coef. p-value ME

RRP 0.758*** (0.002) 0.153 1.021 (0.113) 0.059

0.510* (0.057) 0.109

Hostile -0.317 (0.462) -0.064 -1.043 (0.199) -0.060

-0.723 (0.332) -0.155

Tender -2.339*** (0.000) -0.473 -1.195** (0.042) -0.069

-4.251*** (0.000) -0.910

Diversification 0.797*** (0.001) 0.161 -0.165 (0.676) -0.010

-0.060 (0.792) -0.013

RelativeSize -0.968*** (0.000) -0.196 -0.847 (0.121) -0.049

-0.307 (0.238) -0.066

Bidder lnMV -0.377*** (0.000) -0.076 -0.809*** (0.000) -0.047

-0.141 (0.162) -0.030

Bidder MTBV 0.052*** (0.002) 0.011 0.060 (0.181) 0.003

0.038 (0.109) 0.008

Bidder ROA -2.376*** (0.009) -0.481 -1.664 (0.453) -0.096

-2.797* (0.071) -0.599

Target lnMV 0.600*** (0.000) 0.121 0.696** (0.014) 0.040

0.170 (0.161) 0.036

Target MTBV 0.062** (0.016) 0.013 -0.004 (0.936) 0.000

0.063** (0.049) 0.014

Target ROA -0.154 (0.764) -0.031 0.069 (0.938) 0.004

0.699 (0.219) 0.150

Target Volatility 8.929 (0.254) 1.808 -1.335 (0.952) -0.077

-1.354 (0.866) -0.290

Bidder Volatility 15.449 (0.144) 3.127 35.750 (0.140) 2.068

51.081*** (0.000) 10.936

Bidder Leverage -0.306*** (0.002) -0.062 -0.219 (0.304) -0.013

-0.169 (0.150) -0.036

Target Leverage -0.129* (0.090) -0.026 -0.089 (0.350) -0.005

-0.203** (0.013) -0.043

Bidder CF/E 0.096 (0.791) 0.019 -0.634 (0.359) -0.037

2.044*** (0.001) 0.438

Year Yes Yes Yes

Industry Yes Yes Yes

Constant -2.916** (0.033) -0.057 (0.974) 0.044 (0.976)

N 666 434 633

Pseudo R2 0.345 0.268 0.333

Chapter 4. Tables

220

Table 4.5: The effect of the RRP on offer premiums

This table reports the ordinary least square (OLS) regression results for offer premiums on the RRP, controlling for a series of deal and firm characteristics. Specification (1) reports

the relationship between offer premiums and the reference point effect, specification (2) controls for deal characteristics, specification (3) controls for deal and bidder characteristics,

specification (4) controls for deal and target characteristics, specification (5) controls for all variables. Variable definitions are as in the notes to Table 4.2. Robustness t-statistics are

reported in parentheses. Statistical significance at the 1%, 5% and 10% levels is denoted ***, **and * respectively, is reported alongside coefficients.

Offer Premiums (1) (2) (3) (4) (5) Coef. t-stat. Coef. t-stat. Coef. t-stat. Coef. t-stat. Coef. t-stat.

RRP 0.101*** (5.099) 0.105*** (5.311) 0.110*** (5.575) 0.098*** (4.846) 0.089*** (4.501)

Hostile 0.008 (0.367) 0.002 (0.105) 0.015 (0.698) 0.032 (1.563)

Tender 0.075*** (4.744) 0.074*** (4.695) 0.080*** (5.098) 0.075*** (5.006)

Diversification 0.006 (0.433) 0.004 (0.284) 0.005 (0.369) -0.017 (-1.256)

Stock -0.024 (-1.336) -0.019 (-0.991) -0.026 (-1.392) -0.003 (-0.184)

Cash -0.009 (-0.565) -0.004 (-0.227) -0.019 (-1.123) -0.031* (-1.917)

RelativeSize 0.020 (1.537) 0.021* (1.744) 0.116*** (7.263)

Bidder ROA 0.090 (1.262) 0.019 (0.269)

Bidder MTBV -0.000 (-0.222) -0.000 (-0.010)

Bidder lnMV 0.006 (1.506) 0.056*** (9.627)

Bidder Volatility 0.570 (0.819) 0.768 (1.047)

Target ROA 0.119*** (2.810) 0.141*** (3.347)

Target MTBV -0.001 (-0.806) -0.001 (-0.878)

Target lnMV -0.030*** (-6.512) -0.078*** (-11.703)

Target Volatility -0.035 (-0.065) -0.315 (-0.537)

Constant 0.212*** (3.291) 0.182*** (2.744) 0.116 (1.535) 0.334*** (4.388) 0.150* (1.896)

Year Yes Yes Yes Yes Yes

Industry Yes Yes Yes Yes Yes

N 1878 1878 1878 1878 1878

adj. R2 0.077 0.088 0.089 0.116 0.169

Chapter 4. Tables

221

Table 4.6: The effect of the RRP on offer premiums over time

This table reports the OLS regression results for offer premiums on the RRP by time periods. We divided our sample into

different time periods. Specification (1)-(3) report results before 1990, 1990 to 2000, and after 2000 respectively. Variable

definitions are as in the notes to Table 4.2. Robustness t-statistics are reported in parentheses. Statistical significance at

the 1%, 5% and 10% levels is denoted ***, **and * respectively, is reported alongside coefficients. (1) (2) (3)

Offer Premiums 1985-1989 1990-2000 2001-2014


RRP -0.058 (-0.773) 0.062** (2.392) 0.125*** (3.858)

Hostile -0.049 (-0.463) 0.044 (1.476) 0.031 (1.038)

Tender 0.191*** (2.750) 0.109*** (4.668) 0.048** (2.152)

Diversification -0.021 (-0.395) -0.003 (-0.149) -0.024 (-1.255)

Stock 0.078 (0.931) -0.002 (-0.076) -0.039 (-1.353)

Cash 0.037 (0.564) -0.084*** (-3.258) 0.006 (0.268)

RelativeSize 0.064 (1.413) 0.133*** (5.774) 0.092*** (3.355)

Bidder ROA 0.623** (2.306) -0.010 (-0.104) -0.005 (-0.044)

Bidder MTBV 0.001 (0.273) -0.001 (-0.432) 0.001 (0.848)

Bidder lnMV 0.028 (1.544) 0.058*** (6.615) 0.049*** (5.472)

Bidder Volatility -4.431 (-1.110) 0.534 (0.535) 0.733 (0.611)

Target ROA 0.157 (0.911) 0.188*** (2.996) 0.121** (1.979)

Target MTBV -0.004 (-0.378) -0.000 (-0.040) -0.003 (-1.577)

Target lnMV -0.032 (-1.508) -0.079*** (-7.832) -0.074*** (-7.090)

Target Volatility 1.856 (0.774) -0.330 (-0.415) 0.093 (0.098)

Constant -0.413** (-2.600) 0.087 (0.860) 0.301*** (3.104)

Year Yes Yes Yes

Industry Yes Yes Yes

N 114 934 830

adj. R2 0.190 0.176 0.184

Chapter 4. Tables

222

Table 4.7: The effect of the RRP on offer premiums by valuation periods

This table reports the OLS regression results for offer premiums on the RRP before and after the 2008 financial crisis. The sample was divided into two subsamples: between 2000 and

2007 and between 2008 and 2014, i.e. financial crisis period, presented in specification (4) and (5) respectively. Variable definitions are as in the notes to Table 4.2. Robustness t-

statistics are reported in parentheses. Statistical significance at the 1%, 5% and 10% levels is denoted ***, **and * respectively, is reported alongside coefficients.

(1) (2)

Offer Premiums 2001-2007 2008-2014


RRP 0.094** (2.293) 0.192*** (3.655)

Hostile 0.089** (2.245) -0.003 (-0.073)

Tender 0.017 (0.539) 0.080** (2.514)

Diversification -0.046* (-1.772) 0.010 (0.306)

Stock -0.005 (-0.142) -0.081* (-1.854)

Cash 0.023 (0.784) 0.004 (0.138)

RelativeSize 0.069* (1.892) 0.121*** (2.981)

Bidder ROA -0.083 (-0.694) 0.300 (1.273)

Bidder MTBV 0.001 (0.462) 0.003 (1.059)

Bidder lnMV 0.037*** (3.633) 0.059*** (3.631)


Target ROA 0.151** (2.176) 0.011 (0.097)

Target MTBV 0.001 (0.238) -0.006** (-2.322)

Target lnMV -0.067*** (-5.149) -0.076*** (-4.427)


Constant 0.285** (2.369) 0.102 (0.685)

Year Yes Yes Industry Yes Yes N 496 334 adj. R2 0.190 0.375

Chapter 4. Tables

223

Table 4.8: Univariate analysis by different RRP groups

This table reports univariate analysis results for the offer premium, the bidder and the target’s CAR3 calculated with market model. We divide our sample into those RRP less than 0,

which is the bidder is relatively more undervalued than the target and those RRP larger than 0, which is the bidder is relatively more overvalued than the target. Panel A reports the

univariate analysis results for the offer premium. Panel B reports the univariate analysis results for the bidder CAR3. Panel C reports the univariate analysis results for the target CAR3.

Panel D reports univariate analysis for the offer premium by the method of payment. Here “Cash” represents that acquisitions are 100% financed with cash. “Stock” represents

acquisitions that are 100% financed by stocks. Specifications (3) and (4) report offer premiums of 100% cash-financed acquisitions at RRP>0 and RRP<0 groups. Specifications (3)

and (4) report offer premiums of 100% stock-financed acquisitions at RRP>0 and RRP<0 groups. The mean value, t-statistics, and the number of observations for the offer premium,

the bidder and the target 3-day abnormal returns around the announcement date are reported in each Panel. The mean difference of t-tests is reported at the end of each Panel. T-statistics

are reported in parentheses. Statistical significance at the 1%, 5% and 10% levels is denoted ***, ** and * respectively.

Panel A: Univariate analysis for the offer premium

Offer Premiums t-stat. N

(1) RRP<0 0.279*** (30.20) 723

(2) RRP>0 0.329*** (37.26) 1,155

Mean difference (2)-(1) 0.050*** (-3.79)

Panel B: Univariate analysis for the bidder CAR3

Bidder CAR3mm t-stat. N

(1) RRP<0 -0.006** (-2.27) 723

(2) RRP>0 -0.014*** (-6.37) 1,155

Mean difference (2)-(1) -0.008** (-.2.28)

Panel C: Univariate analysis for the target CAR3

Target CAR3mm t-stat. N

(1) RRP<0 0.191*** (23.81) 723

(2) RRP>0 0.240*** (32.34) 1,155

Mean difference (2)-(1) 0.049*** (4.33)

Panel D: Univariate analysis for the offer premium by the method of payment at different RRP subgroups

(1) (2) (3) (4) Mean difference RRP>0 RRP<0 RRP>0 RRP<0 (1)-(2) (3)-(4) (1)-(3) (2)-(4) Cash Stock

Offer Premiums (%) 35.25*** 26.75*** 31.33*** 29.22*** 8.50*** 2.11 3.92* -2.47

t-stat (25.18) (20.90) (19.07) (14.51) (4.24) (0.80) (1.88) (-1.08)

N 435 291 402 206 724 608 837 497

Chapter 4. Tables

224

Table 4.9: OLS regressions of offer premiums on the RRP: By RRP subsamples

This table reports OLS regression results for the offer premium on RRP. We divide the sample according to the RRP.

Column (1) presents the results for the offer premium on RRP when the bidder is relatively more overvalued than the

target, column (2) presents the results for the offer premium on RRP when the bidder is relatively undervalued than the

target. Variable definitions are as in the notes to Table 4.2. Robustness t-statistics are reported in parentheses. Statistical

significance at the 1%, 5% and 10% levels is denoted ***, **and * respectively, is reported alongside coefficients.

(1) (2)

Offer Premiums RRP>0 RRP<0


RRP 0.164*** (4.766) -0.005 (-0.112)

Hostile 0.003 (0.105) 0.063** (1.968)

Tender 0.090*** (4.317) 0.055*** (2.724)

Diversification 0.009 (0.483) -0.058*** (-2.923)

Stock -0.004 (-0.170) 0.004 (0.159)

Cash -0.024 (-1.080) -0.036 (-1.548)

RelativeSize 0.128*** (5.132) 0.104*** (5.109)

Bidder ROA -0.050 (-0.577) 0.103 (0.881)

Bidder MTBV 0.002 (1.258) -0.002 (-1.129)

Bidder lnMV 0.051*** (6.456) 0.059*** (6.943)

Bidder Volatility -0.763 (-0.750) 2.070* (1.875)

Target ROA 0.168*** (3.439) 0.115 (1.467)

Target MTBV -0.002 (-0.900) 0.000 (0.119)

Target lnMV -0.079*** (-8.750) -0.073*** (-7.066)


Constant 0.060 (0.574) 0.012 (0.078)

Year Yes Yes Industry Yes Yes

N 1155 723

R2 0.220 0.214

Chapter 4. Tables

225

Table 4.10: Do bidder focusing on the RRP overpay for the target?

This table reports the OLS regression and 2SLS regression results for both bidder and target 3-day CARs on offer

premiums. The dependent variable for specifications (1) and (2) is the bidder CAR3 and for specifications (3) and (4) is

the target CAR3, where CAR3 is calculated with the market model. The parameters are estimated in the window [-261,-

28]. Specifications (2) and (4) reports the results of 2SLS regression using the RRP as an instrument variable of offer

premiums. Variable definitions are as in the notes to Table 4.2. Robustness t-statistics are reported in parentheses.

Statistical significance at the 1%, 5% and 10% levels is denoted ***, **and * respectively. The results of the Hausman

test and F-test were reported at the lower part of the table.

(1) (2) (3) (4) Bidder CAR3 Target CAR3

OLS IV OLS IV

Offer Premiums -0.018*** -0.090*** 0.469*** 0.765*** (-2.807) (-2.626) (20.499) (6.597)

Hostile -0.002 -0.002 -0.000 -0.003 (-0.324) (-0.317) (-0.028) (-0.153)

Tender 0.009** 0.017*** 0.041*** 0.010 (2.376) (2.888) (3.102) (0.533)

Diversification 0.003 0.003 0.004 0.004 (0.805) (0.867) (0.375) (0.373)

Stock -0.011** -0.010** -0.045*** -0.045*** (-2.326) (-2.155) (-3.952) (-3.512)

Cash 0.021*** 0.021*** 0.031*** 0.037*** (5.245) (4.555) (2.615) (2.857)

RelativeSize -0.005 -0.005 -0.050*** -0.052*** (-1.116) (-1.311) (-6.136) (-5.505)

Bidder lnMV -0.003*** -0.003*** (-3.916) (-3.230)

Bidder MTBV -0.000 -0.000 (-0.972) (-0.901)

Bidder ROA 0.041** 0.040*** (2.186) (2.998)

Target lnMV -0.002 0.006 (-0.866) (1.447)

Target MTBV -0.002** -0.002 (-2.228) (-1.640)

Target ROA -0.012 -0.017 (-0.469) (-0.800)

Constant 0.015* 0.032*** 0.106*** -0.027 (1.843) (2.706) (5.578) (-0.483)

N 1878 1878 1878 1878

adj. R2 0.060 . 0.374 .

Hausman test X2=4.8934 (p=0.0271) X2=7.9056 (p=0.0050)

F-test X2=60.2985 (p=0.0000) X2=42.9958 (p=0.0000)

Chapter 4. Tables

226

Table 4.11: Do all stock-financed acquisitions driven by the RRP protect the wealth of long-term shareholders?

Panel A of this table reports the univariate analysis results of both the offer premium and the firms’ 36-month market-

adjusted buy-and-hold abnormal returns (BHAR36ma) by RRP quartiles. The sample only consists of 100% stock-

financed acquisitions. Each quartile is assigned a rank from 1 to 4. Rank 1 represents bidders that are relatively more

undervalued than their targets (i.e. RRP<0), and rank 4 represents bidders that are relatively more overvalued than their

targets (i.e. RRP>0). Panel B of this table reports the univariate analysis results of both the offer premium and the firms’

36-month size-adjusted buy-and-hold abnormal returns (BHAR36sa) by the RRP. We divided the RRP into three levels,

the bottom one third or rank 1 refers to bidders are relatively more undervalued, while the top one third or rank 3 refers

to bidders are relatively more overvalued, the middle accounts for the remaining observations. We reported mean value,

t-statistics and the number of the offer premium at each rank. BHAR36ma and BHAR36sa are winsorised at the 1% and

99% level. We performed bootstrap estimation of sampling distribution of BHAR36ma and BHAR36sa at 1000

replications, and report mean value, p-value and the number of BHAR36ma and BHAR36sa of each rank. The mean

difference of t-tests is reported at the end of the table. T-statistics (or p-value) are reported in parentheses. Statistical

significance at the 1%, 5% and 10% levels is denoted ***, ** and * respectively.

Panel A: Market-adjusted BHARs by RRP quartiles

All stock-financed acquisitions Offer Premiums t-stat. N BHAR36ma p-value N

1 (RRP<0) 0.296*** (12.12) 152 -0.348*** (0.000) 137

2 0.251*** (11.74) 152 -0.221** (0.017) 144

3 0.283*** (12.02) 152 -0.198** (0.025) 145

4 (RRP>0) 0.395*** (13.86) 152 -0.161* (0.093) 147

Mean difference 4-1 0.099*** (2.63) 0.187* (0.079)

Panel B: Size-adjusted BHARs by RRP

All stock-financed acquisitions Offer Premiums t-stat. N BHAR36sa p-value N

1 0.294*** (14.53) 201 -0.400*** (0.003) 158

2 0.262*** (13.04) 201 -0.206** (0.040) 180

3 (RRP>0) 0.360*** (15.29) 206 -0.253*** (0.001) 174

Mean difference 3-1 0.066** (2.11) 0.147* (0.073)

Chapter 4. Tables

227

Table 4.12: Endogeneity issues

This table reports the RRP effect on the offer premium by controlling for endogeneity issues. Results from an OLS

regression and a 2SLS regression are presented in this table. The RRP is treated as an endogenous variable, and the market

52 week high, with the bidder and the target 65 week-high treated as instrumental variables. We first obtain the fitted

value from the regression of the RRP on the instrument variables and then replace the RRP with the fitted value. Results

are reported in the “IV” Column. The market 52-week high is defined as the logarithmic term difference between the

highest total market value (CRSP: TOTVAL) over the 335 calendar days ending 30 days prior to the announcement date

and the total market value 30 days prior to the announcement date. The bidder (the target) 65 week-high is 427 calendar

days ending 30 days prior to the announcement date and the stock price 30 days prior to the announcement date. Variable

definitions are as in the notes to Table 4.2. Robustness t-statistics are reported in parentheses. Statistical significance at

the 1%, 5% and 10% levels is denoted ***, ** and * respectively, is reported alongside coefficients. The results of the

Hausman test and the over-identifying restrictions test are reported in the lower part of the table.

Offer Premiums OLS IV


RRP 0.089*** (4.501) 0.085*** (4.853)

Hostile 0.032 (1.563) 0.032 (1.312)

Tender 0.075*** (5.006) 0.075*** (4.518)


Stock -0.003 (-0.184) -0.003 (-0.179)

Cash -0.031* (-1.917) -0.031* (-1.856)

RelativeSize 0.116*** (7.263) 0.116*** (7.719)

Bidder ROA 0.019 (0.269) 0.018 (0.337)

Bidder MTBV -0.000 (-0.010) -0.000 (-0.005)

Bidder lnMV 0.056*** (9.627) 0.056*** (10.282)


Target ROA 0.141*** (3.347) 0.140*** (4.545)

Target MTBV -0.001 (-0.878) -0.001 (-0.883)

Target lnMV -0.078*** (-11.703) -0.078*** (-12.472)


N 1878 1878

adj. R2 0.169 0.169

Hausman test 0.2170 (p=0.6414)

Over-identifying restrictions

(Sargan test) 26.8164 (p=0.0000)

Chapter 4. Tables

228

Table 4.13: Variables correlation matrix

This table reports pairwise Pearson correlation of the variables used in the regression of offer premiums on the RRP. Variable definitions are as in the notes to Table 4.2.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1.RRP 1.0000

2.Hostile -0.0450 1.0000

3.Tender -0.0199 0.1333 1.0000

4.Diversification -0.0072 -0.0175 0.0245 1.0000

5.Stock 0.0957 -0.1122 -0.2918 -0.0289 1.0000

6.Cash -0.0355 0.0518 0.3323 0.1144 -0.5493 1.0000

7.RelativeSize -0.0939 0.2082 -0.0710 -0.1175 0.0101 -0.2716 1.0000

8.Bidder ROA -0.0898 0.0096 0.0884 0.0969 -0.1875 0.1732 -0.1187 1.0000

9.Bidder MTBV 0.0092 -0.0661 -0.0459 -0.0099 0.1511 -0.0770 -0.0789 0.1448 1.0000

10.Bidder lnMV -0.0465 -0.0500 0.0938 0.0764 -0.1840 0.1811 -0.3759 0.3133 0.2044 1.0000

11.Bidder Volatility 0.0667 -0.0353 -0.0688 -0.0681 0.3333 -0.2574 0.1446 -0.4836 0.1110 -0.4176 1.0000

12.Target ROA -0.2785 0.0618 0.0088 0.0286 -0.0651 0.0106 0.1003 0.2997 -0.0042 0.0912 -0.2870 1.0000

13.Target MTBV -0.1282 -0.0369 -0.0187 0.0093 0.1246 -0.0714 -0.0517 0.0379 0.1840 0.1591 0.0662 0.0426 1.0000

14.Target lnMV -0.2091 0.1085 -0.0051 -0.0844 -0.0961 -0.0757 0.1873 0.1755 0.1153 0.6213 -0.3167 0.3045 0.1913 1.0000

15.Target Volatility 0.3084 -0.1088 -0.0072 0.0382 0.2321 -0.0760 -0.1515 -0.2704 0.0929 -0.2373 0.6290 -0.4696 0.0385 -0.5106 1.0000

Chapter 4. Tables

229

Table 4.14: Robustness check: the RRP effect on the probability of using stocks

This table reports logistic regression results of the probability of using stocks on RRP effect. Dependent variable

is “Stock”, which is a dummy variable, taking a value of 1 if acquisitions are 100% financed with stocks, 0

otherwise. Column (1) reports regression results for all-stock financed acquisitions on the RRP for the acquisition

sample that the bidder is relatively more overvalued than the target or RRP>0, column (2) reports regression

results for all-stock financed acquisitions on RRP for the acquisition sample that the bidder is relatively more

undervalued than the target or RRP<0. We control all other variables the same with Table 4.3. There are 50

observations missing due to multicollinearity problem with the year, so that we are left with 1,742 observations,

1,103 for RRP>0 group, and 639 for RRP<0 group. Variable definitions are as in the notes to Table 4.2. P-value

is reported in parentheses. Statistical significance at the 1%, 5% and 10% levels is denoted ***, ** and *

respectively, is reported alongside marginal effects. We transfer coefficients into marginal effect (ME), evaluated

at the sample means of the independent variables.

Binomial Logistic Regression

Stock

(1) (2)

RRP>0 RRP<0

Coef. p-value ME Coef. p-value ME

RRP 0.831*** (0.003) 0.164 -0.232 (0.611) -0.035

Hostile -0.471 (0.235) -0.093 -1.544** (0.019) -0.233

Tender -2.291*** (0.000) -0.453 -3.799*** (0.000) -0.573

Diversification 0.204 (0.237) 0.040 -0.082 (0.752) -0.012

RelativeSize -0.991*** (0.000) -0.196 0.006 (0.982) 0.001

Bidder lnMV -0.431*** (0.000) -0.085 -0.103 (0.319) -0.016

Bidder MTBV 0.036** (0.031) 0.007 0.074*** (0.000) 0.011

Bidder ROA -0.879 (0.273) -0.174 -2.627** (0.026) -0.396

Target lnMV 0.497*** (0.000) 0.098 0.169 (0.162) 0.025

Target MTBV 0.039* (0.090) 0.008 0.087*** (0.003) 0.013

Target ROA 0.144 (0.704) 0.029 -0.192 (0.803) -0.029

Target Volatility -3.052 (0.606) -0.604 14.092 (0.142) 2.124

Bidder Volatility 33.483*** (0.000) 6.621 36.743*** (0.002) 5.538

Bidder Leverage -0.144 (0.124) -0.028 -0.276*** (0.007) -0.042

Target Leverage -0.151** (0.016) -0.030 -0.099 (0.175) -0.015

Bidder CF/E 0.330 (0.338) 0.065 0.342 (0.428) 0.052

Year Yes Yes Industry Yes Yes Constant -1.226 (0.237) -1.744 (0.171)

N 1103 639

Pseudo R2 0.316 0.350

Chapter 4. Tables

230

Table 4.15: Robustness test: The effect of the RRP on the probability of using stocks: Between payment

methods

This table reports multinomial logistic regression results for Stock versus Cash and Stock versus Mixed on the

RRP by different RRP groups. Specifically, RRP>0 represents bidders that are relatively more overvalued than

their targets, and RRP<0 represents bidders that are relatively more undervalued than their targets. “Stock” refers

to acquisitions that are 100% financed by stocks. “Cash” refers to acquisitions that are 100% financed with cash.

“Mixed” refers to acquisitions that are neither 100% cash financed nor 100% stocks financed. There were 5

missing observations that are defined as “Other” in terms of the method of payment in Thomson One. For both

columns, we control for 15 variables as shown in the specification (4) of Table 4.3. Variable definitions are as in

the notes to Table 4.2. P-value is reported in parentheses. Statistical significance at the 1%, 5% and 10% levels is

denoted ***, **, and * respectively.

(1) (2)

Multinomial Logistic Regression Stock Versus Cash Stock Versus Mix

Coef. p-value Coef. p-value

RRP 0.728*** (0.000) 0.425** (0.017)

Hostile -0.943*** (0.006) -0.516 (0.126)

Tender -3.045*** (0.000) -1.831*** (0.000)

Diversification -0.010 (0.947) 0.262* (0.096)

RelativeSize 0.428 (0.143) -0.852*** (0.000)

Bidder lnMV -0.281*** (0.000) -0.228*** (0.000)

Bidder MTBV 0.058*** (0.001) 0.042*** (0.002)

Bidder ROA -1.768** (0.031) -1.294* (0.077)

Target lnMV 0.471*** (0.000) 0.190** (0.015)

Target MTBV 0.059*** (0.003) 0.050*** (0.007)

Target ROA 0.053 (0.887) 0.324 (0.369)

Target Volatility 2.528 (0.651) 4.395 (0.410)

Bidder Volatility 48.423*** (0.000) 19.221** (0.010)

Bidder Leverage -0.181** (0.024) -0.215*** (0.002)

Target Leverage -0.042 (0.451) -0.156*** (0.001)

Bidder CF/E -0.044 (0.885) 0.361 (0.239)

Year -2.097** (0.018) 0.541 (0.494)

Industry Yes Yes Constant Yes Yes N

Pseudo R2

Chapter 4. Tables

231

Table 4.16: Robustness test: The effect of the RRP on offer premiums: By payment methods

This table reports OLS regression results for the offer premium on the RRP. We divided our M&A sample in three

subsamples according to payment methods. Column (1) reports the regression results for the acquisition sample

that is 100% financed by stocks, column (2) reports the regression results for the acquisition sample that is 100%

financed by cash, column (3) reports the regression results for the acquisition sample that is paid with a mixture

of stocks and cash. There are 5 missing observations that are defined as “Other” in terms of the method of payment

in Thomson One. For both columns, we control for 15 variables as shown in the specification (4) of Table 4.3.

Variable definitions are as in the notes to Table 4.2. Robustness t-statistics are reported in parentheses. Statistical

significance at the 1%, 5% and 10% levels is denoted ***, **, and * respectively.

(1) (2) (3)

Offer Premiums Stock Cash Mix


RRP 0.071** (2.239) 0.113*** (3.674) 0.089** (2.303)

Hostile -0.011 (-0.184) 0.038 (1.107) 0.001 (0.030)

Tender 0.073 (1.053) 0.070*** (3.549) 0.098*** (3.434)

Diversification -0.014 (-0.508) -0.009 (-0.467) -0.030 (-1.243)

RelativeSize 0.176*** (4.397) 0.133*** (5.699) 0.097*** (3.389)

Bidder ROA -0.120 (-1.152) 0.303** (2.301) 0.004 (0.034)

Bidder MTBV -0.000 (-0.042) -0.003 (-1.529) 0.003 (1.414)

Bidder lnMV 0.075*** (5.786) 0.038*** (5.027) 0.064*** (4.762)

Bidder Volatility 0.170 (0.138) -0.673 (-0.505) 1.864 (1.392)

Target ROA 0.223*** (3.516) 0.064 (0.940) 0.073 (0.800)

Target MTBV 0.000 (0.145) -0.004 (-1.142) -0.004 (-1.490)

Target lnMV -0.091*** (-6.627) -0.073*** (-7.326) -0.087*** (-6.206)

Target Volatility 0.541 (0.551) 0.026 (0.034) -2.314* (-1.761)

Constant -0.537*** (-3.799) 0.155 (1.322) 0.020 (0.138)

Year Yes Yes Yes Industry Yes Yes Yes N 608 726 539

adj. R2 0.146 0.223 0.218

Chapter 4. Tables

232

Table 4.18: Robustness test: The effect of the RRP on offer premiums: By deal types

This table reports OLS regression results for the offer premium on the RRP. We divided our M&A sample in two

subsamples according to industry relatedness of the two firms involved. Where we define the “Diversification” to

which the primary two Standard Industry Classification (SIC) codes are different between bidders and targets.

Column (1) reports the regression results for the diversified acquisition sample, column (2) reports the regression

results for the undiversified acquisition sample. For both columns, we control for 15 variables as shown in the

specification (4) of Table 4.3. Variable definitions are as in the notes to Table 4.2. Robustness t-statistics are

reported in parentheses. Statistical significance at the 1%, 5% and 10% levels is denoted ***, **, and *

respectively.

(1) (2)

Offer Premiums Diversified deals Undiversified deals


RRP 0.127*** (4.066) 0.067** (2.566)

Hostile 0.023 (0.657) 0.034 (1.254)

Tender 0.062** (2.471) 0.090*** (4.555)

Stock -0.016 (-0.500) -0.002 (-0.081)

Cash -0.035 (-1.259) -0.039* (-1.917)

RelativeSize 0.136*** (4.795) 0.115*** (5.611)

Bidder ROA -0.050 (-0.316) 0.019 (0.236)

Bidder MTBV 0.003 (1.277) -0.001 (-0.620)

Bidder lnMV 0.047*** (5.384) 0.066*** (7.924)

Bidder Volatility -0.658 (-0.479) 1.414 (1.561)

Target ROA 0.136** (2.016) 0.138** (2.472)

Target MTBV -0.002 (-0.657) -0.001 (-0.774)

Target lnMV -0.082*** (-7.785) -0.080*** (-8.938)


Constant 0.046 (0.386) 0.034 (0.327)


N 702 1176

adj. R2 0.159 0.173

Chapter 4. Tables

233

Table 4.18: Robustness test: The effect of the RRP on offer premiums: By deal choice

This table reports OLS regression results for the offer premium on the RRP. We divided our M&A sample in two

subsamples according to deal choice or whether the deal is defined as tender offer according to Thomson One.

Column (1) reports the regression results for the tender offers, column (2) reports the regression results for mergers.

For both columns, we control for 15 variables as shown in the Specification (4) of Table 4.3. Variable definitions

are as in the notes to Table 4.2. Robustness t-statistics are reported in parentheses. Statistical significance at the

1%, 5% and 10% levels is denoted ***, **, and * respectively.

(1) (2)

Offer Premiums Tender Merger


RRP 0.148*** (4.107) 0.078*** (3.464)

Hostile 0.007 (0.190) 0.041 (1.505)

Diversification -0.052** (-2.044) -0.007 (-0.478)

Stock -0.011 (-0.190) 0.005 (0.282)

Cash -0.076** (-2.578) -0.020 (-1.020)

RelativeSize 0.150*** (4.668) 0.108*** (5.862)

Bidder ROA 0.056 (0.489) -0.007 (-0.091)

Bidder MTBV -0.005 (-1.172) 0.001 (0.440)

Bidder lnMV 0.070*** (5.518) 0.052*** (7.675)

Bidder Volatility 4.638*** (3.126) 0.028 (0.033)

Target ROA 0.066 (0.757) 0.167*** (3.512)

Target MTBV -0.002 (-0.800) -0.001 (-0.543)

Target lnMV -0.081*** (-5.372) -0.077*** (-10.246)


Constant 0.325*** (2.764) 0.155 (1.430)


N 380 1498

adj. R2 0.224 0.144

Chapter 4. Tables

234

Table 4.19: Robustness test: The effect of the RRP on offer premiums: By deal completion

This table reports OLS regression results for the offer premium on the RRP. The M&A sample was divided in two

subsamples according to whether the deal is completed during the sample time period studied. Column (1) reports

the regression results for the completed deals or successful deals, column (2) reports the regression results for

deals that are failed to complete or unsuccessful deals. For both columns, we control for 15 variables as shown in

the specification (4) of Table 4.3. Variable definitions are as in the notes to Table 4.2. Robustness t-statistics are

reported in parentheses. Statistical significance at the 1%, 5% and 10% levels is denoted ***, **, and *

respectively.

(1) (2)

Offer Premiums Successful deals Unsuccessful deals


RRP 0.081*** (3.921) 0.119** (2.347)

Hostile 0.021 (0.517) 0.092*** (2.693)

Tender 0.066*** (4.013) 0.100** (2.560)

Diversification -0.015 (-1.007) 0.022 (0.609)

Stock 0.010 (0.515) -0.017 (-0.375)

Cash -0.033* (-1.842) 0.014 (0.311)

RelativeSize 0.130*** (7.074) 0.134*** (3.422)

Bidder ROA 0.024 (0.320) -0.295 (-1.512)

Bidder MTBV -0.001 (-0.441) 0.005 (1.482)

Bidder lnMV 0.052*** (8.597) 0.102*** (4.315)


Target ROA 0.125*** (2.658) 0.291*** (2.830)

Target MTBV -0.001 (-0.534) -0.004 (-0.842)

Target lnMV -0.076*** (-10.891) -0.118*** (-4.747)


Constant 0.022 (0.240) -0.158 (-1.015)

Year Yes Yes Industry Yes Yes N 1597 277

adj. R2 0.164 0.238

Chapter 4. Tables

235

Table 4.20: Robustness test: market-adjusted model

This table reports the OLS regression and 2SLS regression results for both bidder and target 3-day CARs on offer

premiums. The dependent variable for specifications (1) and (2) is the bidder CAR3 and for specifications (3) and

(4) is the target CAR3, where CAR3 is calculated with the market-adjusted model. Specifications (2) and (4)

reports the results of 2SLS regression using RRP as an instrument variable of offer premiums. Variable definitions

are as in the notes to Table 4.2. Robustness t-statistics are reported in parentheses. Statistical significance at the

1%, 5% and 10% levels is denoted ***, **and * respectively. We report the results of the Hausman test and F-test

at the lower part of the table.

(1) (2) (3) (4)

Bidder CAR3 Target CAR3

OLS IV OLS IV

Offer Premiums -0.014** -0.070** 0.469*** 0.688***

(-2.167) (-2.078) (20.616) (6.173)

Hostile -0.002 -0.002 -0.002 -0.004

(-0.385) (-0.366) (-0.127) (-0.218)

Tender 0.008** 0.014** 0.040*** 0.017

(2.102) (2.471) (3.022) (0.970)

Diversification 0.003 0.003 0.003 0.003

(0.764) (0.817) (0.305) (0.312)

Stock -0.010** -0.009** -0.046*** -0.046***

(-2.139) (-2.047) (-4.023) (-3.727)

Cash 0.020*** 0.020*** 0.030*** 0.035***

(5.066) (4.504) (2.581) (2.793)

RelativeSize -0.005 -0.005 -0.051*** -0.053***

(-1.207) (-1.485) (-6.259) (-5.780)

Bidder lnMV -0.003*** -0.003***

(-3.693) (-3.074)

Bidder MTBV -0.000 -0.000

(-0.345) (-0.186)

Bidder ROA 0.034* 0.033**

(1.877) (2.550)

Target lnMV -0.002 0.004

(-0.827) (1.001)

Target MTBV -0.002* -0.002

(-1.757) (-1.372)

Target ROA -0.008 -0.011

(-0.298) (-0.557)

Constant 0.013* 0.027** 0.107*** 0.008

(1.666) (2.297) (5.604) (0.153)

N 1878 1878 1878 1878

adj. R2 0.050 . 0.374 .

Hausman test X2=2.9735 (p=0.0848) X2=4.3522 (p=0.0371)

F-test X2=60.2985 (p=0.0000) X2=42.9958 (p=0.0000)

236

CHAPTER 5: MARKET

ANTICIPATION AND MERGER

DECISIONS

Chapter 5. Market Anticipation and Merger Decisions

237

Abstract

This chapter examines the market anticipation effect on M&A decisions and outcomes. Prospect

theory predicts sequential risk-taking actions by decision-makers when facing losses. This

implication was applied in the M&A context, which is a firm’s major corporate investment decision.

The market anticipates managers making investment decisions according to its risk tolerance, hence

maintaining confidence in the firm (i.e. market anticipation effect). Managers follow this anticipation

to rationalise the M&A motive. The magnitude of losses was measured with the bidder reference

price, which is the distance between the firm’s 52-week high and its current stock price, illustrating

the market perception of a firm’s performance. Investors holding bidders’ stocks whose value is in

significant decline relative to the reference price tend to believe risky projects should be available to

the firm and if managers fail to engage in them, this will destroy the market confidence. Market

anticipation increases managerial incentive to perform well in the face of performance decline and

triggers a risk-taking appetiser. The level of riskiness of investment decisions was measured with the

target firm-specific risk. The principal findings in this chapter are presented as follows: it was found

that the riskiness of managerial decisions is anticipated by the market and this explains the choice of

takeover targets. Managers who make decisions based upon market anticipation will receive positive

market reactions. Market-anticipated risks may expose managers to great shareholder monitoring,

pushing them to work hard to justify the M&A motive. Specifically, managers tend to exhibit great

efforts in deal negotiation, reflected in the lower offer premiums they pay for the target, indicating

that managers are cautious in the face of losses. The quality of their decisions according to the

reference point proves to be value creation, reflected in positive long-term performance. Findings of

this chapter suggest that the reference point effect has predictability for target selection.


238

5.1. Introduction

“Pain, I have already had occasion to observe, is, in almost all cases, a more pungent sensation than

the opposite and correspondent pleasure. The one, almost always, depresses us much more below the

ordinary, or what may be called the natural state of our happiness, than the other ever raises us

above it.”

Adam Smith (1759: 176-177)

One of the most well-documented phenomena of investors’ behaviour is that people tend to take risks

subsequent to follow losses relative to a reference point, predicted in Kahneman and Tversky’s

prospect theory (1979). According to this, the market should be reluctant to realise losses relative to

a reference point, with the anticipation that future gains can be realised through subsequent firms’

risk-taking decisions. In this chapter, this implication is examined in the context of M&As, which is

a firm’s major investment decisions. Market anticipation was measured with the bidder reference

point, the extent to which a bidder’s current price deviates from its 52-week high price. Evidence that

the market tends to gauge gains and losses relative to a salient piece of price information is also

documented in many studies relating to behavioural finance and M&A literature (Odean, 1998;

Kumer et al., 2015; Schneider and Spalt, 2015)75. Risk magnitude of merger and acquisition (M&A)

decisions was measured with target firm-specific risk similar to that employed by Kumar (2009).76

75 Using the firm’s 52-week high as the reference price is not new. Huddart et al. (2009) identified abnormal trading

volumes around the 52-week high. Burghof and Prothmann (2011) and Li and Yu (2012) found that abnormal trading

volumes around the 52-week high reflects investor attitude towards market news and uncertainty information. Baker et

al. (2012) suggest that bidders assess the target with the target 52-week high. Consistently, the reference price was applied

to the bidder. It reflects the extent to which the loss relative to an aspiration level that the market is unwilling to realise,

which triggers expectations of the firm’s risk-taking decision.

76 The construction of target firm-specific risk is presented in the methodology section of this chapter.


239

Whether market anticipation has predictability for corporate investment decisions was investigated,

and the effect of market-anticipated risks on M&A outcomes was examined. The expectation was to

be able to assess whether managers minimise a firm’s value by taking market-anticipated risks.

Explanations regarding why losses trigger managerial willingness of risk-taking are put forward from

the psychological view and provided with empirical evidence relating to the stock market and M&As.

France (1902) explains that people take risks when they encounter losses. He documented a wide

range of historical facts as well as the psychological origin of humans, revealing that the incentive to

take risks is people’s deepest instincts of tasting the joy of life. Further to this, Kahneman and Tversky

(1979) developed a model portraying the relationship between risk and choice, suggesting that people

are loss-aversion and tend to engage in risky decisions in an attempt to compensate losses. Odean

(1998) documents direct evidence in the stock market that investors are reluctant to realise paper

losses, holding the belief that future gains should be generated.

Kumar (2009) suggests investors’ preference for stocks with a lottery feature, which is more

pronounced for those in a lottery-favourable environment and with low incomes. The author noted

that investors are likely to exaggerate a small chance of a large gain. As a result, they choose stocks

with high idiosyncratic volatility, high idiosyncratic skewness and lower stock price (i.e. lottery-like

stocks). Studies suggest that investors who play games of chance are in the hope that their current

wealth will be improved, indicating that investment decisions are motivated by current losses. Barber

and Odean (2008) noted that investors’ trading decisions are based on the focus of their attention.

Both individual and institutional investors will choose eye-grabbing stocks. In this connection, Yu

and Li (2011) and Bhootra and Hur (2013) attribute investors’ attention to the reference point effect


240

that trading strategy is driven by a firm’s peak prices. Similarly, Gamble and Johnson (2014) claimed

that investors gauge their losses and gains relative to prior stock returns and increase their portfolio

risks following losses. Kumar et al. (2015) use prior stock market reactions as reference points and

find that losses reflected in prior stock market reactions drive the firm to take on risky targets.

Many additional studies have also documented salient evidence that losses drive a firm’s risk-taking

decisions. Edmans et al. (2012) suggest current underperformance relative to an aspiration level

increases managerial pressure relative to their peers which pushes the manager to take on risky

projects. In a similar vein, Morrow et al. (2007) reported that managers feel great pressure when their

firm’s current performance is far below the market expectations, leading their attempts at value

creation through M&A activities. Kim et al. (2011) documented evidence of desperate managers

whose aim is to grow their firm through M&As when the firm’s performance falls greatly below the

market expectations. Gorton et al. (2009) proposed a theory of mergers relating to firm size, arguing

that firms whose current performance is poor are associated with great pressures, which motivates

managers to start a race for firm size through M&A activities to eliminate threats of being taken

over.77

The magnitude of loss of the firm was measured with the bidder reference price, i.e. a price deviation

of a firm’s current price from its highest stock price occuring in a one-year window preceding the

takeover announcement date. This gives the market insight into how the firm is currently performing

relative to the 52-week high, which is a salient reference price that shapes investors’ minds regarding

77 In support of this, Dickerson et al. (2003) noted that attack, through acquisitions, is the best way to eliminate takeover

threats.


241

change in a firm’s profitability. Thus, this measure reflects the magnitude of loss of a firm. Investors

may lose confidence in the firm when this loss is large, inducing their managers to take on risky

projects that could fundamentally change their current wealth status, or blame them for not working

hard. Following Kumar (2009) and Schneider and Splat (2015), the riskiness of the M&A decisions

was measured with target firm-specific risks, which are more easily to be assessed by the market.

Large firm-level risks also expose managers to great shareholder monitoring, motivating the firm to

work hard.

Prior studies have provided supportive evidence as to why a firm tends to make decisions based on a

single piece of information though in truth that they have a full access to the information about their

own firms. Brandenburger and Polak (1996) suggest that the manager is concerned about the market

and tend to make decisions that the market wants to see. Thus, it should be suggested that managers

should make M&A decision according to the market anticipation. Sacheti et al. (2016) argued that

the manager whose decisions follow market anticipation is attempting to avoid the market’s short-

term criticisms. Similarly, Kau et al. (2008) found that firms listen to the market, in that they cancel

investment decisions if there are negative market reactions.78

Using M&As as a platform to explore the reference point effect on major corporate investment

decisions has three main reasons. First, M&A outcomes will alter a firm’s status and change the

wealth of shareholders. Brenner (1983) indicates that a primary motive of risk-taking is to

fundamentally change the relative position in the distribution of wealth. A firm will be blamed for

78 Luo (2005) and Chikh and Filbien (2011), applying this in different markets, found a consistent view with that of Kau

et al. (2008).


242

not making sufficient effort for their shareholders when it fails to undertake the investments the

market expects. Secondly, the fact that the shareholder motivates the firm to increase risks is common

in the CEOs’ M&A compensation literature. For instance, Grinstein and Hribar (2004) find risk-

taking managers are compensated with lucrative compensation packages. Graham et al. (2013)

suggest that shareholders are likely to hire managers with risk-taking incentives to avoid costly

incentive pay. M&As allow investigation of whether firms’ major investment decisions made upon

market anticipation receive market recognition. Third, the researcher further explored whether

managers can control risks through skills and efforts, which allows observers to distinguish risk-

taking from stock market gambling, which purely depends on probability (Kumar, 2009).

This chapter has two main motivations. First, it provides novel insights into the debate on managerial

risk-taking incentives. Previous studies have found that risk-taking incentives are driven by

managerial overconfidence or agency problems and yield contradictory results with respect to bidder

performance. Whilst Malmendiar and Tate (2008) claimed that overconfident managers who tolerate

large risks tend to destroy the wealth of shareholders, Croci and Petmezas (2015) indicated that

managerial risk-taking incentive creates value for the firm. Another line of literature links managerial

risk-taking attitudes with CEOs’ M&A compensation. Datta et al. (2001) indicate that high pay-

performance creates managerial value-maximising incentives. However, this view is challenged by

Sanders and Hambrick (2007) who suggest that an increase in risk-taking does not necessarily

enhance firm value. This research investigates the risk-taking motive with reference point theory,

which explains M&A decisions from human nature response (France, 1902). It was expected that a

firm will follow the market anticipation following losses, and managers should rationalise their risk-


243

taking incentive to avoid the market criticism. Thus, market perception has a substantial impact on

M&A decisions.

Secondly, only a few papers link the reference point to target risks. For example, Kumar et al. (2015)

use stock market reaction to a prior acquisition as the reference point. The authors find that prior

market reaction loss induces a firm to acquire risky targets. However, how the decision of prior

acquisition influences the current M&A decision remains largely controversial (Billett and Qian, 2008;

Aktas et al., 2009)79. Moreover, Kumar et al. employed target stock volatility as a proxy for M&A

riskiness. However, it measures the total risk of the market and the firm, making it hard to distinguish

the effect of firm-specific risk from the effect of the market-wide risk.80 To combat these issues, in

this chapter, the bidder 52-week high is used as the reference point, which is a more direct measure

for current market reaction. Schneider and Spalt (2015), who applied Kumar’s idea (2009) of

employing the measure for stock riskiness to the target firm, found a negative relationship between

target risks and bidder performance. They ascribe this to a managerial gambling preference. However,

Schneider and Spalt fail to answer whether such risk-taking has a rational motive. If the market

believes that there are risky projects available to a firm while managers do not take any actions, this

will lead to the investors’ belief that managers have a quiet life by entrenching themselves and

destroying the firm’s value (Bertrand and Mullainathan, 2003). On the other hand, when M&A

decisions are made in relative to market anticipation, the market tends to create an image that the firm

79 Aktas et al. (2009) found the learning hypothesis that managers tend to adjust their behaviour according to the previous

acquisition while Billett and Qian (2008) suggest the self-attribution hypothesis that initial success generates

overconfident managers who tend to engage in more acquisitions, which generates lower shareholder wealth. 80 Often the market tends to react more significantly to the firm-level news than the market news. Firm-specific risks can

be driven by any news relating to the firm level. For example, CEO turnovers, or earnings announcement, which are more

sensitive to outside investors.


244

is strong, or the firm strives hard to protect shareholders’ wealth, thus enhancing its confidence in the

firm’s prospects. Croci and Petmezas (2015) suggest that managers with risk-taking incentives tend

to select better quality acquisitions. In addition, March and Shapira (1987) indicate that risk-taking is

an essential part of the managerial role. Based on these arguments, bidder performance is expected to

improve when risk-taking is undertaken with market consent.

Using a sample of 2,018 U.S. public acquisitions announced between 1985 and 2014, a positive

relationship between riskiness of M&A decision and the bidder reference point was established,

suggesting an important role of played by market anticipation in corporate investment decisions. The

results of this chapter continue to hold after a series of tests. Further, it was found that bidders’

announcement returns increase with market-anticipated risks, implying that the market will push the

manager to work hard. Further investigation shows that managers who are exposed to observable

risks tend to exhibit greater efforts on deal negotiation and attempting to signal to the market that

they are good managers, which is consistent with Harford et al. (2012) and Bertrand and Mullainathan

(2003).81

Several contributions have been made to previous literature. This is the first paper, to our knowledge,

investigates the reference point effect on managerial risk-taking incentives in the context of M&As.

It is shown that humans’ psychological bias has a direct impact on corporate investment decisions.

This study focuses on the M&A motives of firms who have announced an M&A deal, which allows

making the distinction between the performance of bidders who undertake M&As with market

81 Their studies find that entrenched managers tend to have safe lives and attempt to avoid being monitored, which leads

to overpayment. It is expected that managers making risky decisions will signal to the market that the interests of the

manager and the shareholder are closely linked.


245

anticipation and those without, eliminating any concerns that excessive risks are due to managerial

overconfidence and agency problems. It is expected that those M&A motives are followed by

negative market reactions because the market does not anticipate the firm to do so. This chapter

contributes to behavioural finance literature by directly documenting that not only individual

investors but also institutional investors are influenced by the reference point effect while making

decisions in uncertainty. Further, it offers new insights to the existing literature regarding managerial

risk-taking incentives. It was found that market-anticipated risks have a significant impact on M&A

decisions, attributable to one of the important sources of bidder performance improvement.

The remainder of this chapter is organised as follows: Section 2 summaries relevant literature. Section

3 designs hypotheses. Section 4 summarises the data and presents methodologies. Section 5 examines

the reference point effect on M&A decisions, and the relationship between market-anticipated risks

and bidder performance. Section 6 conducts a series of robustness checks. Section 7 concludes the

chapter.


5.2.1. Reference point theory and gambling

France (1902) documents a wide range of interesting facts relating to the history of gambling across

the world and provides experimental evidence explaining the motivation of people engaging in

gambling activities. He concludes that it is human nature to get involved in games of chance and

gambling. People have a gambling spirit and believe in luck. With regard to the human gambling

spirit, France affirms that people tend not to leave the table while a big risky game is in progress. The

human gambling spirit implies that humans naturally explore uncertain events and their attention


246

tends to be intense when information is highly uncertain. He explained luck with the fact that

gamblers believe that if they lose in the first half of the game, they will eventually win back in the

second half. This implies that people tend to align games of chance with an exaggerated feeling for

their skills. Such a stimulated mental intention drives people not to think wisely but to rely on the

gambling impulse.

Kahneman and Tversky (1979) developed prospect theory with an S-shape model portraying the

relationship between risk and choice. It suggests that people seek risk when the current status is below

their aspiration level. They are more sensitive to feelings about loss than gains given that the slope in

the loss domain is much steeper than that in the gain. Due to the fact that people are loss-averse, they

have a great incentive to compensate their losses through risk-taking decisions. The model ascribes

the reason for gambling to people’s strong tendency to loss-aversion.

Kumar (2009), following this line of literature, investigates the stock market and finds evidence that

stock-purchasing decisions of individual investors are loss driven. According to the author, individual

investors have a tendency to buy lottery-type stocks. The lottery-type stocks are those with a

combination of high idiosyncratic volatility and skewness, and a low price. When idiosyncratic

volatility is high, investors may overestimate the chance of extreme events occurred in the past would

appear in the future. When idiosyncratic skewness is high, investors may overestimate the probability

of making huge gains by investing less. The lower price reduces the costs of participation, giving

investors an illusion of control. When a panel data set, collected from a large U.S. brokerage house,

was analysed containing trading positions of individual investors between 1991 and 1996, the author

revealed that investors’ preference for lottery-type stocks increased when economic conditions


247

became worse or their income was low. Their data also show that low-income investors underperform

the market, indicating that gambling preference is associated with worse performance. Their study

reflects investors’ gambling spirit in investors’ trading strategy in the stock market.

However, it may be argued that the reason for underperformance is not simply the investors’ risk-

taking behaviour since the investors with a lower income could also be less experienced individuals

who are not able to manage the risk of their portfolios in the stock market. In this case, they might

take too many unexpected risks. Borna and Lowery (1987) define this as ‘pure gambling’ in which

people have no control over the outcome of the events but rely heavily on the probability of the

outcome. These authors noted another form of gambling where people’s skills have an impact on the

outcome of gambling,82 implying that acquisitions belong to this form of gambling since managers

with superior skills can better manage the risks.

Rather than Kumar (2009), who investigated the gambling spirit of less experienced investors in the

stock market, Schneider and Spalt (2015) focused on the motivation of gambling for institutional

investors undertaking M&A activities. Following Kumar’s work (2009), Schneider and Spalt

measured target riskiness with its idiosyncratic volatility, which captures the overall risk of the M&A

decisions. Using a sample of public U.S. M&As between 1987 and 2008, the authors found that risky

firms are likely to be acquired, showing the strong gambling preference of bidder managers. Further,

they suggest that bidders who acquire risky targets destroy the wealth of shareholders and the value

of firm reflected in both short and long runs, concluding that the propensity for managerial gambling

82 For example, Borna and Lowery (1987) further illustrate that one’s good knowledge of horse may result in successful

horse-race betting, while rolling a dice is pure gambling in which depends on probability.


248

leads to value-destroying deals. However, their results can be interpreted that managers who gamble

by means of M&As do not offer the market a rational motive given the lack of the current status of

the bidder. Based on this argument, it cannot be determined whether a firm listens to the market and

acquire targets accordingly. Investors might falsely believe that the M&A motive is simply because

their managers ‘go with their guts’ and do not coincide with the market reaction, thus bidders receive

negative market reactions.

5.2.2. Reference point theory and risk-taking

Kahneman and Tversky (1979) suggest that people are aggressive when facing losses, as they believe

there is a chance to break even, whilst people are risk-averse when the possible outcome of the

decision is positive. March and Shapira (1987) provided supportive evidence that managers tend to

avoid losses when their performance meets or exceeds market expectations. Their survey summarises

many studies that link managerial risk perspectives and market expectations. It reveals that some

managers interviewed believe that risk-taking is in an essential role of managers and the satisfaction

of success is obtained from the degree of risks taken. While experiencing a decline in performance,

managerial risk-taking incentive tends to be greater. A majority of managers, according to their

surveyed study, believe that their skills can reduce risks.

Studies also provide empirical evidence supporting the view that managers take risks when facing

performance decline. Morrow et al. (2007) investigated firms’ strategies in the situation where a firm

is unable to meet market expectations. Their study suggests that managers whose firms currently

underperform the market are associated with mounting pressures from their investors whose aim is

to achieve the aspired performance levels. This motivates managers to take actions in an attempt to


249

turn around the table. Morrow et al.’s data show that acquisitions can create value for the firm,

reasoning that the market expects that the firms to manage risks with better management skills in

integrating existing resources with new resources. They also noted that firms facing a decline in

performance may decline further when taking actions that have no effect on performance, implying

that the firm may perform even worse in the face of a decline in performance when it fails to persuade

the market.

Likewise, Kim et al. (2011) found managers are subject to the pressure of firm’s growth when facing

a decline in stock market performance. The decline in performance indicates a lack of growth for the

firm, which motivates the firm to gain organisational recovery via M&As. By doing so, managers

tend to pay high offer premiums to achieve the deal. Their study interprets their results with reference

point theory of M&A. Kim et al. indicate the driving force of overpayment is because of firms’

tendency to evaluate their current performance relative to their historical performance and their peers’

performance. When they currently perform below these performance levels, they become risk-takers,

paying high offer premiums. However, firms that currently perform well will be aligned with market

expectation that the firm will hit a similar level of growth in the future, and thus managers are not

willing to engage in acquisitions that often destroy the wealth of shareholders.

Edmans et al. (2012) also assert that firms that engage in M&As is because they are under great

pressure from their industry peers. Gorton et al. (2009) proposed a theory of mergers that links the

M&A motive with takeover threat, suggesting a firm that makes a takeover decisions does so to

eliminate takeover threats. This finding is appealing for medium-size firms of the industry in their

sample. The primary reason for firms engaging in unprofitable acquisitions is to eliminate takeover


250

threats and thus maintain control of the firm. If a firm’s primary motive is to complete takeovers to

eliminate any pressure or takeover threats from the same industry, they are likely to choose takeovers

that are easier to complete and firms that potentially bring forth higher returns. Following this

argument, Ferreira and Laux (2007) found firms with larger idiosyncratic risk have fewer anti-

takeover defences, thus, implying that firms that lack of takeover resistance are suitable takeover

targets. In addition, those firms have higher levels of private information flow and information about

future earnings in stock prices, which increase attractiveness for those under great pressures.

Earlier studies indicate that firms’ salient historical prices are often reference point prices that the

market uses to assess the firms’ performances. Baker et al. (2012) proposed the target 52-week high

as a reference point price in M&A pricing decisions. Target shareholders are less likely to forsake

control of the firm when an offer price received is significantly lower than the reference point price.

They judge their managerial performance during M&As according to their expectations, implying

that those failing to meet their performance expectations will incur criticism and even face the risk of

shareholder litigation.

Li and Yu (2013) attributed their findings that investors make decisions based on the salient news of

the market to investors’ attention hypothesis, proposing two proxies for investors’ attentions: the

market 52-week high and the historical high. Specifically, the current market price relative to the

market 52-week high is used as a proxy for investors’ under-reaction on the assumption that sporadic

news leads conservative investors to under-react either to good or bad news as they are not fully

incorporated in the market. Whilst the current market price relative to the historical high is used as a

proxy for investors’ over-reaction, reasoning that a series of good or bad news accumulate great deal


251

of market attention, investors who make decisions according to the market historical high are over-

react. Li and Yu’s data cover the average index of the Dow Jones between 1928 and 2009 and define

the market 52-week high as a ratio of the current Dow index and its 52-week high, and the historical

high as a ratio of the current Dow index and its historical high. Their findings show that investors

whose attention is based on the historical high are associated with negative market returns in the

future, whilst those whose attention is based on the 52-week high are associated with positive market

returns, in that both over and under-reactions are eventually corrected by the market. Their study

suggests that investors are attention-driven, making stock market predictable.

Bhootra and Hur (2013) suggest the recent effect, indicating that investors tend to have much more

attention on the events occurred in the recent past than in the distant past. They proposed a proxy for

investors’ recent effect based on the timing of the 52-week high, i.e. the number of days since the

date of the 52-week high arrived. Bhootra and Hur’ s study documents the evidence of investors’

attention-driven performance, indicating that firms whose 52-week high occurred in the recent past

outperform those whose 52-week high occurred in the distant past. In particular, Bhootra and Hur’ s

findings show that the top 10% of the stocks whose 52-week high occurred in the recent past

outperform the bottom 10% of the stocks whose 52-week high arrived in the further distance in

monthly average returns by 0.7%. Investors create a strong momentum of profitability for the stocks

that have achieved their best performance recently. On average, profits created by investors’ attention

are twice as large for stocks that attain peaks recently than a long time previously. On the whole, their

study indicates that recent events are likely to associate much more of investors’ emphasis, implying

that a firm whose performance has recently reached its peak might less likely to initiate a takeover

for the reason that takeovers destroy the wealth of the bidder shareholders.


252

Many additional studies also suggest the important role of the reference point effect on risk-taking

actions and employ various reference point candidates. Kumar et al. (2015) used the stock market

reactions to a previous acquisition as a reference point for the decision of the current acquisition. The

authors proposed two competing hypotheses relating to the relationship between risk-taking and

choices to examine the decision-making process: reference point effect and the house money effect.

The former indicates that people take risks when facing losses, while the latter suggests that people

take risks when facing gains. Analysing a sample of 823 acquisitions between 1990 and 2006, Kumar

et al. found that firms increase M&A riskiness with both the abnormal dollar gains from the takeover

announcement of a prior acquisition and the abnormal dollar losses, suggesting both theories play a

role in risk-taking. Their data show that loss-aversion driven risk-taking is associated with weak

negative market reactions to the current acquisition, while risk-taking followed by gains has an

insignificant impact on M&A performance. Their study documents evidence that managers take on

risky projects according to market reactions. However, their study measures target riskiness with

target stock return volatility, which includes both the market overall risk and firm-specific risk. This

proxy is problematic, as it cannot be distinguished whether risk-taking is driven by firm-specific or

market overall risks. If the market overall risk is large, investors may believe that risk-taking is

necessary and engaging any projects may as well associate with large risks, making investors feel no

difference to M&As.

Meanwhile, Gamble and Johnson (2014) found evidence suggesting that investors increase their risk

tolerance following losses. Their study focused on individual investors’ trading strategy in the stock

market. Their sample consists of 78,000 households between 1991 and 1996, which were collected

from a large discount brokerage firm in the United States. Investors tend to exit the market when they


253

experience large gains or losses in the first six months of the year, whilst those remaining in the

market tend to lower (increase) their risks subsequently follow gains (losses), which is explained by

the disposition effect.

5.2.3. Risk-taking and M&A outcomes

The relationship between managerial risk-taking attitude and value creation is far less clear according

to March and Shapira (1987) who affirm that risk attitudes vary significantly among different

individual managers. Earlier M&A studies provide rather unambiguous results regarding the

relationship between the firm’s risk-taking attitudes and M&A performance. A thread of literature

suggests that a firm’s risk-taking incentive and M&A outcomes are negatively related, stating the

reason for this is agency conflicts between the managers and the shareholders (Grinstein and Hribar,

2004; Sanders and Hambrick, 2007; Graham et al., 2012) or managerial overconfidence (Malmendier

and Tate, 2005; 2008).

Graham et al. (2012) focused on the relationship between managerial risk-taking attitudes and

corporate investment actions and measured a CEO’s risk profile with a series of survey questions.

The authors established that approximately 10% of the top managers interviewed are risk-averse,

whilst a majority of managers display considerable risk-taking attitude and are likely to engage in

value-destroying deals. Moreover, due to the high level of risk tolerance, managers are optimistic,

leading to a great use of short-term debt to finance M&As. Their study also suggests that the firm is

more willing to recruit risk-taking managers on behalf for their decision-making, as recruiting risk-

averse managers make compensations that motivates costly managerial risk-taking incentive. These

findings imply that it is that shareholders expect their managers to be risk-taking but not too optimistic.


254

Likewise, Grinstein and Hribar (2004), who investigated the compensation effect on managerial risk-

taking incentive in a sample of 327 M&A deals between 1993 and 1999, noted that 39% of companies

reward their managers who complete the deal. This leads the researcher to believe that the managerial

incentive of value-maximising tends to reduce if the structure of compensation83 motivates the

managers to take reductant risks. It is also noted that managers in pursuit of high compensations

overpay for the targets (Shleifer and Vishny, 1988).

Malmendier and Tate (2008) found overconfident managers are more likely to engage in risky

acquisitions than their non-overconfident counterparts. Their M&A sample consists of 477 large

publicly traded U.S. firms of Forbes 500 from 1980 to 1994. The study measured CEO

overconfidence with option-based measures, showing that overconfident bidders lower the wealth of

shareholders for 75 basis points as opposed to rational bidders. The authors indicate that compared

with non-overconfident CEOs, overconfident CEOs are less likely to create value via M&As.

Specifically, they are more likely to conduct diversifying acquisitions associated with negative

abnormal returns, as overconfident managers overestimate their skills in an unfamiliar area of

industry. In addition, overconfident managers tend to engage in more acquisitions, indicating that

they underestimate the risks of acquisitions.

Other literature finds contrasting evidence that a managerial risk-taking attitude leads to positive

M&A outcomes. Croci and Petmezas (2015) studied the relationship between CEO risk perspectives

and M&A decisions in a sample of 2,056 bidders engaged in 9,789 acquisitions over the period 1997

83 In this regard, Datta et al. (2001) found that managers of high equity-based compensation firms tend to take value-

enhancing M&A projects. Their study shows compensation motivates risk-taking managers to create value for the firm.


255

to 2011 and measured managerial risk perspectives with vega and delta. In their study, vega refers to

CEO wealth to stock return volatility while delta refers to CEO wealth to stock price following the

studies of Guay (1999) and Core and Guay (2002). Their findings also show that bidder

announcement returns increase with vega at the 5% significance level, which suggests little evidence

that risks driven by overconfidence destroy the wealth of shareholders. Rather, managers with great

risk-taking incentive enhance the wealth of shareholders. The authors conclude that managers with a

high level of risk-taking tend to select quality deals. Likewise, Gervais et al. (2011) found that

overconfident managers have a high-risk profile, making shareholders’ convexity of the pay-

performance incentive redundant. Their study suggests that overconfidence serves as a solution that

reduces managerial risk aversion, thus saving the firm’s investment costs. According to their study,

shareholders expect their managers to be risk-taking, while the firm performs significantly lower than

their level of expectation and it will reward managerial risk-taking behaviour with compensation. As

such, shareholders might believe that their interests are aligned with the managers if they take on

risky projects on their behalf.

Moeller (2015), who investigated approximately 25,000 companies over a 20-year period that

involved in over 265,000 M&A transactions, also found a positive relationship between risky

strategies and a firm’s performance. He noted that ‘M&A is now a well-exploited strategy. The

opportunities have therefore shrunk and risk-free deals are unlikely to yield the desired results. In

the current environment, portfolio managers therefore have to take risks.’ This indicates that, in

current investment environment, managers can hardly perform well without engaging in risky M&As.


256

Harford et al. (2012) found that risk-averse managers generate negative market reactions. Studying a

sample of 1,905 M&As made by U.S. public bidders between 1990 and 2005, the authors find that

entrenched managers who avoid being monitored by block holders tend to use cash instead of stocks

to finance M&As, and moreover they select public targets and overpay for the takeover targets. Their

study compares risk-averse managers to those who make value-destroying deals for fear of losing

their jobs. Bertrand and Mullainathan (2003) proposed the quiet life hypothesis, whereby entrenched

managers are risk-averse.

Many additional studies have attributed the reason for risk-taking actions driving a firm’s positive

performance to the fact that managers are able to take risks that the market anticipates, as the market

tends also to be risk-taking when facing losses. Managers do what the market anticipates can achieve

a better performance, since it creates an insight in the market that the firm is confident in turning

around its performance, and it is more importantly, the interests of the managers and the shareholders

are closely aligned. Certain literature suggests the listening propensity of managers who make

decisions according to market reactions. Brandenburger and Polak (1996) asserted that managers tend

to make decisions that the market thinks it is right rather than their belief that it is for the interests of

the firm, reasoning that ‘the stock market has opinions as to what choices firms should make’. In this

case, managers make decisions according to the market reactions rather than private information they

have. Sacheti et al. (2016) illustrated this with a cricket game by investigating how cricket players

make decisions when facing social pressure. The cricket captains do not make optimal decisions that

can maximise the probability of winning when they are subject to external pressure. Similarly, Sacheti

et al. suggest that managers who make decisions based on the market anticipation are attempting to

avoid criticisms by the market.


257

Following this trains of ideas in relevant literature, studies document evidence that managers make

decisions in favour of the market anticipation (Luo, 2005; Kau et al., 2008; Chikh and Filbien, 2011),

as managers only complete the deal associated with positive market reactions. It is generally believed

that a firm is a financing contract between the managers and shareholders, in that the managers are

hired by shareholders, and should make major investment decisions that are highly committed to

completing the transactions. Managers who serve the best interests of their shareholders should award

by the market. On the other hand, managers should exhibit sufficient effort to prove they are good

managers, thus secure their jobs (Lehn and Zhao, 2006). When a firm experiences losses in

performance, it should become more aggressively to persuade the market that they can reverse the

situation. Risky deals attract great deal of market attention, pushing the firm to work diligently.

Therefore, it is expected that one of the positive sources of abnormal returns generated from M&As,

based on the rationale that the manager and the market have a common belief to the firm’s prospects.

This chapter uses the bidder reference point as a proxy for the market anticipation, which is a

deivation of the bidder’s current price from its 52-week high, reflecting the change in market reaction

relative to the bidder’s best performance achieved over the past year prior to the takeover

annoucement. If bidder’s current price is significantly discounted from the reference point price,

shareholders suffer mental losses, which motivate the managers to take more risks. On the other hand,

the role of the reference point effect rationalises managerial risk-taking behaviour. Managers who

conduct M&As according to the level of risks that the market anticipates leads the market to believe

that their interests are closely aligned with those of the managers, as managers are also exposed to

the strict shareholders’ monitoring by taking on risky projects. While the market reacts negatively to

those taking unexpected risks, such as managerial overconfidence or agency problems. It is also


258

suggested that the market shows negative reactions to those remaining salient (i.e. entrenched

managers) when they are expected to take an action. This researcher measures riskiness of M&As

with target firm-specific risks in line with the Schneider and Spalt (2015). This research also focuses

on the performance of the firm whose actions are in favour of its shareholders. Not only is the short-

term performance investigated but also the long-term, since it should be possible to observe the

quality of decisions over a longer period of time.

5.3. Hypotheses development

Tversky and Kahneman (1974) indicate a silent piece of information shapes people’s minds and is

used as a reference point for decision-making. Kahneman and Tversky (1979) applied their idea to

the loss-averse tendency, suggesting that mental losses are gauged by the reference point. People

become aggressive when their expectation deviates greatly from the reference point. Odean (1998)

put forward prospect theory in the stock market , documenting evidence that investors are reluctant

to realise loss, thus they retain losers too long in the belief that future gains can be realised through

risky projects. Based on this argument, shareholders holding the firm’s stocks expect the current

losses to be turned round by managers. This author predicts that market reactions to a firm’s current

status relative to the time when the 52-week high arrives are perceived as a loss, which requires the

firm to respond to the market that the firm is striving to achieve profitability. It is generally believed

that M&As are the easiest solution to boost a firm’s performance when it faces a decline in

performance as noted in the work of Morrow et al. (2007) and Kim et al. (2011), however, M&As,

on average, do not benefit bidder shareholders, based on the assumptions that managers take too many

unexpected risks that destroy the wealth of their shareholders (Malmentier and Tate, 2008), or

manager avoid taking expected risks for job security (Bertrand and Mullainathan, 2003; Harford et


259

al., 2012). By contrast, when managers are cautious about the M&A motive and the take risks as the

market anticipates, they reveal to the market that there is no misalignment of interests between the

shareholders and the managers. The reference point price reflects the market’s risk-taking profile,

giving managers an insight into how many risks the firm should take to rationalise its M&A motive.

This author measures the loss magnitude of M&A decisions with target firm-specific risks and expect

that they are increased with perceived loss measured by the bidder reference point. Based on the

reference point effect, this research suggests that there exists a level of risks the market is willing for

the firm to take. When the market feels loss, the firm is anticipated to gain by taking on risky projects.

This leads the first testable hypothesis of this chapter,

H1: There is a positive relationship between the bidder reference point and target risks.

The firm under pressures is prone to listen to the market (Luo, 2005; Kau et al., 2008). When engaging

in risky projects, managers show that they are working hard for the shareholders instead of enjoying

safe lives when the firm’s performance is declining (Bertrand and Mullainathan, 2003). Their risky

actions also attract greater market attention, leading to great managerial incentive to making profits

for the shareholders. Barber and Odean (2008) suggest the attention-grabbing hypothesis that

investors tend to buy eye-catching stocks. France (1902) explains this phenomenon as the human

nature to resolve uncertainty. For instance, people may find it difficult to leave the table when risky

bets are placed unless the outcome appears. Therefore, it should be expected that managers so as to

prove themselves as “good managers” will take levels of risks that the market can tolerate and manage

the deal afterwards to gain market recognition.


260

H2: Market-anticipated risks are positively associated with bidder announcement returns.

According to Morrow et al. (2007) and Kim et al. (2011), M&A deals are an important way of value

enhancing for managers when the firm experiences a decline to performance. The pressure triggers

managerial risk-taking incentive. In the meanwhile, managers work hard in an attempt to calm down

the market’s panic. They should judge how many risks should be taken based on the market

anticipation measured with the bidder reference point. Unlike overconfident managers who overpay

for the takeover targets, those taking anticipated risks creates the market an image of smart and

diligent managers. By substantiating this market belief, managers reduce offer premiums and

integrate the firm’s resources following a merger. Therefore, this leads the last testable hypothesis of

this chapter that,

H3: Managers making decisions according to market-anticipated risks work hard.


5.4.1. Data

A sample of 36,506 U.S. public acquisitions announced between Jan 1, 1985 and December 31, 2014

was collected from Thomson One. This database contains all deal information used in this chapter.

Deals defined as repurchases according to Thomson One were excluded, which was left with a sample

of 13,482 acquisitions. Acquisitions are material decisions to bidders only if their outcomes can

change shareholders’ fundamental wealth. Based on this criteria, bidders were required to acquire at

least 50% of the target shares during the transactions, which yields 11,043 acquisitions. The deal

sample was matched with COMPUSTAT and CRSP. The former database contains the firm’s


261

accounting information and the latter one includes all stock price information. Both bidder and target

stock price studied were not a missing value in CRSP, as they are used to calculate target idiosyncratic

volatility, and bidder announcement returns, which yields 5,640 acquisitions. The method of payment

information is not missing, resulting in 5,297 acquisitions. All accounting variables used in the

regressions were required to be available at the fiscal year end prior to the announcement date in

COMPUSTAT. Both bidder and target market value (MV), return-on-asset (ROA), market-to-book

value (M/B), leverage measured with debt-to-equity ratio (D/E) and relative size measured with deal

value divided by the bidder MV, were not with a missing value. It was left with 2,278 acquisitions

after excluding all bidder variables with a missing value, and result in a final sample of 2,018

acquisitions after excluding all target variables with a missing value.

The reference point effect on M&A decisions was studied by controlling a number of deal and bidder

characteristics. Bidder size is expected to be positively related to target risks, as larger bidders tend

to have greater ability to absorb risks. Firm’s growth opportunities were measured with M/B. It is

generally believed that greater opportunities are associated with higher risks. Firm’s profitability was

measured with ROA. It should be expected that the lower ROA bidders tend to be more aggressive

and take more risks than their higher counterparts. Leverage was measured with D/E, reflecting risks

in capital structure. Lower D/E firms are capable in taking more risks than their higher counterparts

(Uysal, 2011). Small targets with lower profitability and growth opportunities tend to be risky, since

they may need more of bidders’ efforts in integration process. At the deal level, the methods of

payment, deal types and deal attitude that have a significant impact on M&A decisions addressed in

the M&A literature were taken into consideration (Croci and Petmezas, 2015; Schneider and Spalt,

2015).


262

A summary statistics for the M&A deal sample was presented in Table 5.1. Of 2,018 acquisitions,

773 were all-cash financed acquisitions, 632 were all-stock financed acquisitions, and 581 were

mixed acquisitions (financed with a mixture of cash and stocks).84 There was a relative smaller

proportion of tender offers as opposed to mergers: 530 and 1,488, hostile acquisitions (129) and

diversified acquisitions (661), and there is a larger proportion of successful acquisitions as compared

with unsuccessful acquisitions: 1593 and 398.

A summary statistics for variables used in regressions was presented in Table 5.2, with variable

definitions are as in the note of this table. As seen, bidders are generally larger, with higher

profitability and greater growth opportunities than targets. In public acquisitions, bidders are those

with larger size and greater power than targets, so that they are capable of managing the deals

following a merger. These findings are consistent with previous M&A studies (Moeller et al., 2004;

Rau and Vermaelen, 1998).

In Table 5.3, the whole sample was divided into risky and less risky acquisition subsamples according

to the riskiness level of the target firm, the definition of which is by Kumar (2009). The mean, median

and statistical differences between the two subsamples was presented in this table. Kumar (2009)

suggests that firms with a lower price and higher firm-specific risks display higher levels of risks,

increasing the riskiness of M&A decisions (Schneider and Spalt, 2015). Similarly, risky acquisitions

refer to those with a target whose idiosyncratic volatility is above the median value of the sample,

and whose price is below the median value of the sample (i.e. target price), while less risky

84 The method of payment for 32 acquisitions is defined as “other” according to Thomson One.


263

acquisitions refer to the opposite. As seen, bidders engaging in risky acquisitions have a larger

deviation of the current price from the 52-week high price than those engaging in less risky

acquisitions: 0.423 to 0.172, implying that firms perceived as losses are likely to increase risks in the

major investments. Risky acquisitions involve a smaller bidder with lower ROA, growth

opportunities, and leverage. One possible explanation for this is that small bidders are more

vulnerable, which push managers to eliminate threats through risk-taking activity (Gorton et al.,

2009). The statistics also reveal that bidders in the risky acquisition subsample are likely to control

risks by choosing a smaller target than those in the less risky acquisition subsample, based on the

rationale that those smaller targets are less overvalued.

5.4.2. Methodology

5.4.2.1. Target risks and the bidder 52-week high

Target idiosyncratic volatility was measured with standard deviation of residuals computed from a

Fama and French 5-factor model using daily returns from the period of 6 months to 1 month prior to

the announcement date. Fama-French 5-factor returns were collected from the French’s website.

Compared with a 3-factor model, the 5-factor model controls for additional profitability and

investment factors that have an substantial impact on major investment decisions (Fama and French,

2015),85 which is presented as follows:

1 2 3 4 5( )i f m f iR R R R SMB HML RMW CMA (5.1)

85 Schneider and Spalt (2015) used a Fama-French 3-factor model, whereas Kumer (2009) used the 4-factor model to

obtain idiosyncratic volatility. The results obtained in this chapter are similar when calculating target idiosyncratic

volatility with these two models.


264

where iR relates to the individual firm returns, fR relates to risk-free returns, relates to returns on

value-weighted (VW) market portfolio, SMB HML RMW CMA relate to returns on diversified

portfolios of small stocks minus big stocks, high B/M stocks minus low B/M stocks, strong

profitability stocks minus weak profitability stocks, and stocks of low investment firms minus stocks

of high investment firms, respectively. i relates to the residuals of the model.

Target idiosyncratic volatility was then calculated with the standard deviation of i , denotes

iiv in

equation (2).

2

1

n

i

iiiv

N

(5.2)

Finally, target risks are a dummy variable, taking value of 1 if the target is above the median value of

its idiosyncratic volatility and below the median value of its price ending 30 days prior to the

announcement date, whereas the price was expressed in a logarithmic term. Kumar (2009) indicates

stocks with high idiosyncratic volatility indicate that extreme returns occurred in the past are highly

likely occur again, and “cheap bets” are more attractive to investors with the risk-taking incentives.

Therefore, M&A decisions are risky if the target displays a higher level of idiosyncratic volatility and

a lower level of price.

The main variable of interest to be assessed is the bidder reference point (BRP), which is constructed

as follows:


265

, , 30ln 52 lni i t i tBRP weekhighprice Stockprice (5.3)

where the iBRP is defined as logarithmic term differences between the bidder highest stock price over

335 calendar days ending 30 days prior to the announcement date and the price ending 30 days prior

to the announcement date. Scaling the bidder price on 30 days prior to takeover announcement

eliminates any concerns regarding price information leakage around the announcement date biases

the actual stock price.

5.4.2.2. Short-term method

Bidder announcement returns were calculated with the market model, with parameters were estimated

over a period 261 up to 28 trading days prior to the takeover announcement date and a 3-day event

window is used.

1it mt itR R (5.4)

where itR denotes holding period returns (CRSP: RET) for firm i in the period t,

mtR denotes value-

weighted market returns including dividends (CRSP: VWRETD), it denotes the error term.

5.4.2.3. Long-term method

Following Loughran and Vijh (1997) , the firm’s long-term performance was calculated with the

market-adjusted buy-and-hold abnormal returns (BHARs). It should be expected that the outcome of

decisions to be only assessed in the long term. A firm’s 36-month BHARs were calculated with the

following equation:


266

,

1 1

(1 ) (1 )T T

it it index t

t t

BHAR R R

(5.5)

where itR denotes arithmetic returns for firm i on day t. ,index tR denotes the arithmetic return for the

market index on day t. Bootstrapping test was employed to deal with skewness, and the details of this

test was presented in the methodology section of the Chapter 4.


Multivariate analysis was used to examine the relevant factors that have impacts on bidder

announcement returns (Draper and Paudyal, 2008), which is presented as follows:

( 1, 1) 1

1

N

i i i i

i

CAR BRP X

(5.6)

( 1, 36) 1

1

N

i i i i

i

BHAR BRP X

(5.7)

where ( 1, 1)CAR relate to bidder cumulative returns of a day prior to and a day after the announcement

date, ( 1, 36)BHAR refer to market-adjusted buy-and-hold abnormal returns up to 36 months following

acquisitions. The main variable to be assessed is the BRP. iX relate to a series of control variables

that have a significant impact on bidder announcement returns in M&A literature (Moeller et al.,

2004; Alexandridis et al., 2013).


267


5.5.1. Does market anticipation affect M&A decisions?

Table 5.4 presents the logistic regression model of the market anticipation on managerial investment

decisions. The results this researcher obtained show a positive relationship between the BRP and

target risks. Specifically, specification (5) shows that the BRP coefficient is positive and significant

at 1% level (coefficient 1.259, p = 0.000). This coefficient was translated into marginal effect, and

presented alongside Specification (5), yielding similar interpretation of OLS regression. It is therefore

interpreted as that for every 10% increase in the BRP increases a 2.2% likelihood of the firm choosing

a risky target rather than a less risky target. The signs of control variables are generally consistent

with the earlier predictions of this chapter. The goodness-of-fit tests show that logistic regression

model has no specification errors (p = 0.890). 86 The principal results continue to hold when

accounting for different sets of control variables, including deal, bidder and target characteristics, as

reported from specifications (1) to (4) respectively, and all regressions control for year and industry

effects.87

The results for specification (5) indicate that the firm makes major corporate investment decisions

according to market-anticipated risks, which is interpreted with the reference point theory of M&As.

Specifically, the market tends to treat the firm 52-week high price as bidder’s potential profitability

and feel mental losses when its current price is significantly deviated from this aspiration level. This

increases their risk tolerance and pushes the firm to take risks. According to what the prospect theory

predicts (Shefrin and Statman, 1985; Odean, 1998), investors holding losers would presumably

86 The null hypothesis posits that the model has no specification error. The test results lead us not to reject the null

hypothesis.

87 Our results are also consistent when controlling for firm-fixed effect.


268

believe that the firm can generate future profitability via embarking on risky projects. If the firm

responds to the market correctly, the market confirms the belief that the firm works in favour of

shareholders’ interests. On the other hand, managers are also subject to take more risks especially the

firm’s performance is temporarily depressed (Kim et al., 2013). The reference point enables the firm

to justify their risk-taking behaviour without bearing market’s blame (Sacheti et al., 2016). The firm

will also take risks to maintain the confidence of shareholders. The firm persuades shareholders that

current losses will be realised through decisions the market anticipates. Findings can be also

interpreted as the disciplinary effect of takeover: acquiring risky targets will also expose managers to

shareholder monitoring, which threatens managers’ job security when they perform below the

market’s anticipation (Lehn and Zhao, 2006). By following the market’s anticipation, managers can

release pressure from making major investment decisions.

5.5.2. Do market-anticipated risks create value?

Thus far, the author has examined that M&A decisions are influenced by what the market has

anticipated. In this section, there follows an investigation of the bidder performance whether the firm

takes risks that the market anticipates. There are a variety of sources driving the manager to take

unanticipated risks, such as managerial overconfidence (Malmendier and Tate, 2008) or agency

problems (Grinstein and Hribar, 2004), which may mask the true effect of risks on bidders’ M&A

performance. It is expected that if the firm that takes risks with a rational motive will be rewarded by

the market. So as to disentangle the effect of expected risks from the unexpected risks, an interaction

variable using target risks and the bidder reference point was contructed, with the expectation that

targets risks positively related to bidders’ announcement returns on the condition that the firm takes

risks according to what the market anticipates.


269

Table 5.5 presents OLS regressions of bidder announcement returns. Specifications (1) and (2) show

the effects of target risks and the bidder reference point on bidder announcement returns, respectively.

It emerged that either target risks or the BRP alone does not have an impact on bidder announcement

returns. While specification (3) shows a significantly positive relationship between the joint effect of

target risks and the bidder reference point on bidder announcement returns, suggesting that target

risks can predict bidder’s performance when a bidder’s decision is made according to the market

anticipation. In addition, the results indicate that either target risks or the bidder reference point tend

to decrease bidder announcement returns. On the whole, the findings are consistent with the

prediction that firms taking risks observed by the market will receive positive market reactions.

5.5.3. What is the source of bidder performance improvement?

Since the manager is also exposed to the strict shareholders’ monitoring when taking risks that the

market can observe. Managers will attempt to prove themselves as “good managers” in order to avoid

market disappointment (Luo, 2005). According to this, it is believed that managers wish their efforts

to be easily seen by the market. In this case, there is an expectation that managers will exhibit great

efforts on deal negotiation, reflected in offer premiums they pay for the target, in that the offer

premium determines the wealth distribution of the shareholders.

Table 5.6 presents results of OLS regressions of offer premiums. As seen, specification (3) of this

table shows that both target risks and the bidder reference point will increase offer premiums. Bidders

may find targets with greater uncertainty are hard to value or currently underperformed bidders have

less bargaining power and pay hefty offer premiums to target shareholders accordingly. However, the

joint effect of target risks and the bidder reference point decreases offer premiums significantly


270

(coefficient -0.0856, p = 0.000). These findings are consistent with our hypothesis suggesting that

managers will have a direct incentive to decrease offer premiums to show negotiation skills or efforts

to the market. The findings of this table suggest that the primary M&A motive of managers to deal

with pressures of a decline in performance is to content their shareholders.

5.5.4. Do market-anticipated risks push the firm to work hard?

It cannot be expected that the market can fully understand the firm’s decisions in the short run, as

uncertainty prevents the market from assessing the quality of decisions until they are gradually

resolved in the long term. This leads the research to investigate the market-anticipated risks effect on

the firm’s long-term performance.

Results of Table 5.7 report that either target risks or the bidder reference point is positively associated

with the long-term performance, presented in specifications (1) and (2). It should be expected that

target risks are influenced by the bidder reference point as the firm makes decisions based on the

market. As such, an interactive term between target risks and the bidder reference point was

constructed, and presented in specification (3). Results show that effects of these two independent

variables on bidder’s long-term performance diminish when including an interaction variable

between target risks and the bidder reference point. The joint effect of target risks and the bidder

reference point increases with bidder’s long-term performance. These findings suggest that each

individual factor partially explains bidder’s long-term performance and has a substantial impact on

each other. The firm gains positive market reactions in the long term when the firm enhances the

confidence of shareholders. Furthermore, it is suggested that managers will be rewarded when their

efforts are recognised.


271


5.6.1. Various reference point candidates based on firm level.

The robustness checks in this chapter analysed the reference point effect on target risks using a series

of other reference points addressed in previous literature, as presented in Panel A of Table 5.8. First,

the firm’s historical high was used as a reference point, with the expectation that bidder historical

high plays a role in firm’s corporate investment decision-making. A firm’s historical high reflects

investors’ attentions over a longer period of time relative to the 52-week high, and is regarded as a

reference point candidate for conservative investors (Li and Yu, 2012). Results in specification (1)

show a positive relationship between bidder historical high and target risks, suggesting that

conservatism investors also expect the firm to take risks when it faces a decline in performance.

Secondly, a recent ratio was employed as a proxy for investors’ attention, which measures risk-taking

timing for the market and the measure is similar to Bhootra and Hur (2013). The measure is

constructed as 1 minus the number of days since the firm’s 52-week highest stock price arrives

divided by 365 days. Larger of the ratio indicates that the 52-week high arrives more recently. It

should be expected that events occurred in the recent past are more attractive to the market than

occurred in the far distant. If this is the case, the firm should be reluctant to embark on risky

acquisitions when the firm’s best performance recently arrives than in the far distant, since the best

performance arrives more recently will likely to compensate market’s feelings about loss and increase

the attitude of risk aversion. Our results are consistent with the prediction and presented in

specification (2). Thirdly, the prior market reaction was used as a reference point, taking value of 1

if prior market reaction is negative, and 0 if it is positive. If managerial risk-taking is driven by the

market, current market reaction to losses will induce the firm to take on risky projects. Results in

specification (3) show that prior market losses trigger risk-taking behaviour. Finally, it is expected


272

that managers are more risk averse than their shareholders as Kumar (2009) suggested. Using prior

ROA relative to current ROA of a firm as a reference point, taking value of 1 if current ROA is lower

than that of the last year, 0 otherwise. It became apparent that ROA has no predictability for risk-

taking, as presented in specification (4). It can be interpreted as follows: managers are insensitive to

current fundamental loss; instead, they make decisions heavily upon market reactions. Overall, it was

found that the bidder 52-week high has the greatest predictability for risk-taking after controlling for

all relevant reference point candidates, as presented in specification (5).

Results of Table 5.9 show other reference points that influence managerial risk-taking incentives.

First, the market 52-week high was used as a reference point candidate. Baker et al. (2012) suggest

that the market returns are an important influential factor for the firm 52-week high. The firm’s overall

risk-profile changes according to the market condition. If the market risk is high, both the market and

the firm may believe that risk-taking is granted. Results of specification (1) show that the market 52-

week high is positively related to target risks, at 10% significance level. Secondly, peer pressure was

used as a proxy, which reflects the firms’ pressure from their industry peers. An increase in peers’

pressure increases a firm’s risk-taking incentive. According to Edmans et al. (2012), managers whose

firms currently underperform, compared with their peers in the same industry, tend to have

considerable pressures that trigger them to take more risks. Following this, a measurement of peer

pressure was constructed using the logarithmic term differences between the bidder 52-week high

and the 52-week high of the industry medians, depicting the extent to which a firm’s performance

below that of the industry’s median, taking value of 1 if the firm performs below the industry 52-

week high median, 0 otherwise. It is expected that managers feel pressures from their industry peers

when their firms perform significantly lower than the industry median. As seen, specification (2)


273

shows a positive relationship between peer pressure and target risks. Once again, the significance and

sign of the bidder 52-week high does not change when including both the market 52-week high and

peer pressure in specification (3), suggesting that the bidder 52-week high is a competent reference

point candidate being used to anticipate the firm’s following investment decisions.

5.6.2. Subsample tests

The results are robust in different subsamples by the methods of payment, deal types, whether the

deal is diversified and whether the deal is successful, as presented in Table 5.10. However, it was

noted that the reference point effect is not pronounced among all-stock financed acquisition and

diversified acquisition subsamples. One possible explanation for this is that all-stock financed

acquisitions have those financially constrained bidders whose investment opportunities are limited

(Eckbo et al., 2016), while acquiring diversified targets may create obstacles for shareholders, as it is

hard for the market to judge whether the firm is able to manage in an unfamiliar area of industry.

5.6.3. Alternative proxies for target risks

Following Schneider and Spalt (2015), M&A riskiness was also measured with the target

idiosyncratic risk and conducted with an OLS regression of target idiosyncratic risk on the bidder

reference point. Results of specification (1) of Table 5.11 measures target risks with total volatility,

which is employed by Kumar et al. (2015). Specification (2) measures target risks whose residuals

were obtained from the market model. Specification (3) measures target risks whose residuals were

obtained from Fama-French 4-factor model, which is consistent with the measure of Kumar (2009).

It was evident that there is a positive relationship between the bidder reference point and target risks.


274

In addition, the residuals obtained from the market model and the 4-factor model do not vary

significantly when used for target risks.

5.6.4. Risk-taking and M&A performance

Results that the market-anticipated risks taken by the firm increase bidders’ short- and long-term

abnormal returns continue to hold by different estimation models or window periods selected. Table

5.12 reports the relationship between bidder announcement returns and the market-anticipated risks.

The first three specifications measure bidder announcement returns with the market model and uses

a 5-day window while the last three specifications measure bidder announcement returns with the

market-adjusted model with a 3-day window. A positive relationship was found between bidder

announcement returns and the market-anticipated risks, as reflected in specifications (3) and (6)

respectively. Once again, the results show that the firm takes risks according to the market’s

anticipation gain rewards. Table 5.13 shows the relationship between the long-term performance and

the market-anticipated risks. 12- and 24-month windows were used as robustness checks for the 36-

month window of BHARs. The results obtained in this table show that the market-anticipated risks

are positively related with bidders’ long-term performance, which are consistent with the results

reported in Table 5.7.

5.7. Conclusion

In this chapter, this researcher investigates how market anticipation influences firms’ M&A decisions,

and the effect of market-anticipated risks on M&A outcomes. It became evident that perceived loss

measured by the extent to which the deviation of the current price to the firm 52-week high price

triggers managerial risk-taking incentives. Market-anticipated risks lead to positive bidder


275

announcement returns, which can be interpreted as managerial great efforts on deal negotiation.

Managers who take risks according to the market risk perspective will increase the managerial

incentive of proving themselves good decision-makers. By doing so, managers work smart and

diligent, reflected in lower offer premiums they pay for the targets and positive market reactions in

the long term.

This chapter contributes to managerial risk-taking incentives and behavioural finance literature. It

was found that the risk-taking preference of individual investors can be transferred to that of

institutional investors through M&As. The market anticipates firm’s next move, and the firm’s market

decision should be rationalised by shareholders. It became apparent that the market anticipation is a

practice for the firm to follow in order to gain positive market reactions. In addition, though it is hard

to explain whether risk-taking incentive gains market recognition, managers taking risks according

to the market anticipation serves one of the important sources for positive market reactions.

Chapter 5. Tables

276


This table reports summary statistics for 2018 U.S. domestic public acquisitions announced between 1985 and 2014. The

number N denotes the number of deals per year. The third and fourth columns present mean and median of deal value.

The fifth to the seventh columns present the method of payment. Here “Stock” refers to all-stock acquisitions. “Cash”

refers to all-cash acquisitions. “Mix” refers to acquisitions that neither all-stocks nor all-cash acquisitions. “Tender” refers

to tender offers. “Diversified” refers to diversified deals in which the primary two Standard Industry Classification codes

are different between bidders and targets. “Hostile” refers to hostile deals. “Successful” refers to whether deals are

completed during the sample period, there are 1991 deals with information about deal status.

Year N Deal Value ($ mil) Payment Methods Tender Diversified Hostile Successful

Mean Median Cash Stock Mix Yes No Yes No Yes No Yes No

1985 15 325.85 30.80 6 3 2 9 6 2 13 1 14 6 3

1986 21 225.92 61.20 14 2 3 12 9 9 12 1 20 15 3

1987 33 690.41 66.00 15 7 10 12 21 14 19 5 28 27 4

1988 40 986.62 116.10 18 11 10 20 20 12 28 6 34 23 15

1989 38 209.58 86.95 21 11 6 11 27 14 24 1 37 26 8

1990 29 481.63 21.60 15 8 5 11 18 9 20 3 26 20 8

1991 37 102.29 26.82 9 18 4 13 24 15 22 - 37 26 8

1992 21 136.98 40.26 6 12 2 5 16 8 13 2 19 14 6

1993 45 515.02 93.60 14 16 14 10 35 18 27 5 40 30 14

1994 52 260.96 86.95 13 30 8 10 42 14 38 7 45 37 14

1995 92 572.47 88.67 17 54 19 15 77 36 56 6 86 73 16

1996 89 726.33 157.54 19 44 26 20 69 30 59 6 83 74 15

1997 133 633.15 217.88 23 60 50 24 109 46 87 3 130 111 22

1998 141 1186.73 140.99 35 56 49 31 110 44 97 3 138 126 15

1999 163 1505.82 304.21 58 58 46 45 118 67 96 14 149 131 32

2000 137 2388.55 352.84 30 69 37 31 106 49 88 6 131 113 24

2001 115 1052.84 111.58 35 42 38 26 89 39 76 5 110 96 19

2002 52 1665.40 222.40 22 13 17 14 38 17 35 4 48 47 5

2003 77 755.08 126.23 32 17 28 23 54 19 58 5 72 67 10

2004 72 2678.95 438.18 29 17 26 9 63 22 50 3 69 64 8

2005 75 2535.13 327.64 36 13 26 16 59 26 49 5 70 61 14

2006 88 2444.81 509.29 49 14 25 16 72 31 57 4 84 72 16

2007 67 1319.47 686.22 44 6 17 21 46 17 50 2 65 51 16

2008 63 1990.04 230.27 39 7 16 24 39 13 50 10 53 39 24

2009 53 2912.25 268.89 18 13 20 19 34 17 36 - 53 47 6

2010 61 1719.53 440.71 35 9 14 18 43 14 47 5 56 47 14

2011 48 2334.53 546.40 21 8 18 13 35 12 36 9 39 28 20

2012 43 1258.88 606.27 29 3 11 12 31 19 24 - 43 40 3

2013 54 1703.16 781.04 38 3 11 20 34 13 41 3 51 38 16

2014 64 5140.02 1224.41 33 8 23 20 44 15 49 5 59 44 20

Total 2018 1488.04 210.11 773 632 581 530 1488 661 1357 129 1889 1593 398

Chapter 5. Tables

277


This table presents the numbers, means, medians, and standard deviations of variables. Target idiosyncratic volatility is

defined as the standard deviation of the residuals computed from a Fama-French 5 factor model using daily returns from

the period of 6 months to 1 month prior to the announcement date. Target price is target stock price ending 30 days prior

to the announcement date, expressed in logarithmic term. Bidder 52-week high is defined as logarithmic term difference

between the bidder highest stock price over 335 calendar days ending 30 days prior to the announcement date and the

price ending 30 days prior to the announcement date. Bidder 3-day CARs are bidder cumulative abnormal returns 3 days

around the announcement date calculated with market model, whose parameters are estimated using the period of 261

days to 28 days prior to the announcement date. BHAR36ma refers to 36-month market-adjusted buy-and-hold abnormal

returns. Deal characteristics include stock, cash, diversification, hostile, and tender offer which are dummy variables,

taking value of 1 if acquisitions are 100% financed with stocks, 100% financed with cash, diversified merger, hostile

merger, and tender offer, and 0 otherwise. Deal value is raw value taken from Thomson One. “Relative Size” is defined

as deal value divided by bidder MV. Firm characteristics include both bidder and target firm characteristics. MV is defined

as the product of market price and outstanding shares (CRSP: SHROUT*PRC), expressed in logarithmic form. MTBV is

market-to-book value, defined as market equity to book equity, where book equity is total shareholders’ equity (Compustat:

SEQ) plus deferred taxes and investment tax credit (Compustat: TXDITC) minus the preferred stock redemption value

(Compustat: PSTKRV). ROA is return-on-asset ratio, defined as net income (Compustat: NI) divided by total asset

(Compustat: AT). Leverage is measured by debt-to-equity ratio, defined as total long-term debt (Compustat: DITT)

divided by book equity. Other variables are those mainly used in robustness checks. Offer premiums are defined as

logarithmic term difference between offer price and target price ending 30 days prior to the announcement date. Bidder

historical high is defined as logarithmic term difference between bidder highest stock price ending 30 days prior to the

announcement date and the price ending 30 days prior to the announcement date. Recency ratio is defined as 1 minus the

number of days divided by 365 days. Market 52-week high is defined as the logarithmic term difference between the

highest total market value (CRSP: TOTVAL) over the 335 calendar days ending 30 days prior to the announcement date

and the total market value 30 days prior to the announcement date. Peer pressure is defined as the difference between the

bidder 52-week high and the bidder 52-week high of the industry median. Prior market loss is a dummy variable, taking

value of 1 if the firm’s market returns calculated from 365 calendar days before announcement date to 28 calendar days

before takeover announcement date is positive, 0 otherwise. Prior ROA loss is a dummy variable, taking value of 1 if the

firm’s current year’s ROA is less than the past year, 0 otherwise. All accounting variables were at the fiscal year prior to

the announcement date, and continuous variables are winsorised at the 1% and 99% level.

Chapter 5. Tables

278


Variable Mean Median N

Main variables

Target idiosyncratic volatility 0.040 0.033 2018

Target stock price 2.253 2.409 2018

Bidder 52-week high 0.308 0.165 2018

Bidder CAR3 -0.005 -0.005 2018

Bidder BHAR36m -0.075 -0.159 1867

Offer premiums 0.304 0.286 1869

Deal characteristics

Deal value 1488.040 210.113 2018

Cash 0.383 - 2018

Stock 0.313 - 2018

Hostile 0.064 - 2018

Tender 0.263 - 2018

Diversification 0.328 - 2018

Relative size 0.429 0.214 2018

Bidder characteristics

Bidder MV 7.227 7.200 2018

Bidder M/B 3.989 2.536 2018

Bidder ROA 0.018 0.046 2018

Bidder leverage 0.638 0.314 2018

Target characteristics

Target MV 5.173 5.080 2018

Target M/B 2.649 1.767 2018

Target ROA -0.053 0.024 2018

Target leverage 0.622 0.151 2018

Other variables

Bidder historical high 0.919 0.711 2018

Bidder recency ratio 0.564 0.613 2018

Market 52-week high 0.062 0.025 2018

Peer pressure 0.114 -0.022 2018

Prior market loss 0.379 - 2018

Prior ROA loss 0.457 - 2018

Chapter 5. Tables

279

Table 5.3: Risky and less risky acquisition sample

This table presents summary statistics for two subsamples by the median value of target idiosyncratic volatility and target

price ending 30 days prior to the announcement date, expressed in logarithmic term. “Risky acquisitions” refer to targets

that are above the median value of their idiosyncratic volatility and below the median value of their stock price, whereas

“Less risky acquisitions” indicate the opposite. Means and medians of the two subsamples are reported, the number of

observations are reported at the lower part of the table. Median values of dummy variables are omitted as they do not

have any statistical meaning. Variable definitions are as in the notes of Table 5.2. Statistical tests for differences in means,

and medians between the two samples are presented. P-value is reported in parentheses. Statistical significance at the 1%

level, 5% level, 10% level, denoted ***, **, and * respectively.

Variables

(1)

Risky acquisitions

(2)

Less risky acquisitions

Difference (1)-(2)

Mean Median N Mean Median N Mean p-value Median p-value

Bidder 52-week high 0.465 0.308 746 0.216 0.120 1272 0.000 0.000

Bidder CAR3 -0.004 -0.007 746 -0.006 -0.004 1272 0.553 0.765

Bidder BHAR36ma -0.069 -0.280 699 -0.079 -0.114 1168 0.813 0.001

Offer premiums 0.383 0.400 661 0.262 0.258 1208 0.000 0.000

Bidder size 6.132 5.896 746 7.870 7.824 1272 0.000 0.000

Bidder M/B 3.581 2.450 746 4.228 2.607 1272 0.027 0.001

Bidder ROA -0.025 0.039 746 0.043 0.049 1272 0.000 0.000

Bidder leverage 0.506 0.173 746 0.714 0.398 1272 0.001 0.000

Target size 3.609 3.580 746 6.090 5.986 1272 0.000 0.000

Target M/B 1.912 1.229 746 3.082 2.076 1272 0.000 0.000

Target ROA -0.168 -0.045 746 0.015 0.040 1272 0.000 0.000

Target leverage 0.523 0.043 746 0.679 0.241 1272 0.039 0.000

Hostile 0.047 - 746 0.074 - 1272 0.017 -

Tender 0.252 - 746 0.269 - 1272 0.406 -

Diversification 0.362 - 746 0.307 - 1272 0.012 -

Stock 0.389 - 746 0.269 - 1272 0.000 -

Cash 0.355 - 746 0.399 - 1272 0.049 -

Relative size 0.372 0.182 746 0.462 0.234 1272 0.001 0.000

Chapter 5. Tables

280

Table 5.4: Reference point effect on M&A decisions

This table presents results of logistic regressions of target risk on bidder 52-week high. Dependent variable is a

dummy variable, taking value of 1 if targets that are above the median value of their idiosyncratic volatility and

below the median value of their stock price ending 30 days prior to the announcement date. Specifications (1) to

(4) present results with different sets of variables, including deal, bidder and target characteristics respectively,

and specification (5) controls for all of these variables. All regressions include year and industry effect whose

coefficients are omitted. Variable definitions are as in the notes of Table 5.2. P-value is reported in parentheses,

marginal effect for specification (5) is reported alongside coefficients. Goodness-of-fit test is reported at the lower

part of the table. Statistical significance at the 1% level, 5% level, 10% level, denoted ***, **, and * respectively.

Dep. var. Target risks (1) (2) (3) (4) (5)

Bidder 52-week high 1.602*** 1.673*** 1.273*** 1.237*** 1.259*** 0.223 (0.000) (0.000) (0.000) (0.000) (0.000)

Hostile -0.350 -0.221 0.171 0.183 0.032 (0.122) (0.372) (0.567) (0.540)

Tender -0.149 -0.428*** -0.031 0.021 0.004 (0.273) (0.007) (0.865) (0.907)

Diversification 0.225** 0.584*** 0.251 0.175 0.031 (0.045) (0.000) (0.116) (0.292)

Stock 0.066 -0.248 -0.020 0.030 0.005 (0.643) (0.116) (0.920) (0.885)

Cash 0.115 0.009 -0.525*** -0.553*** -0.098 (0.430) (0.958) (0.008) (0.006)

Relative size -0.313*** -1.365*** -0.053 0.150 0.027 (0.004) (0.000) (0.748) (0.461)

Bidder MV -0.559*** 0.107* 0.019 (0.000) (0.062)

Bidder M/B -0.011 -0.007 -0.001 (0.253) (0.528)

Bidder ROA -0.768* -0.034 -0.006 (0.093) (0.954)

Bidder leverage 0.027 -0.006 -0.001 (0.617) (0.937)

Target MV -1.406*** -1.501*** -0.266 (0.000) (0.000)

Target M/B -0.045** -0.043** -0.008 (0.023) (0.031)

Target ROA -2.376*** -2.290*** -0.406 (0.000) (0.000)

Target leverage 0.046 0.046 0.008 (0.356) (0.352)

Year & Industry Y Y Y Y Y

Constant -2.496*** -2.441*** 2.550*** 5.997*** 5.698***

(0.000) (0.000) (0.000) (0.000) (0.000)

N 2018 2018 2018 2018 2018

Pseudo R2 0.162 0.171 0.301 0.517 0.518

Pearson chi (2) X2=1885.56 (p = 0.890)

Chapter 5. Tables

281

Table 5.5: OLS regressions of bidder CARs on market-anticipated risks

This table presents results of OLS regressions of bidder announcement returns on target risk. We include an

interaction variable between target risk and bidder 52-week high in specification (3). Variable definitions are as

in the notes of Table 5.2. P-value is reported in parentheses. Statistical significance at the 1% level, 5% level, and

10% level, denoted ***, ** and * respectively.

Bidder CAR3 (1) (2) (3)

Target risks -0.0024 -0.0148*** (0.553) (0.003)

Bidder 52-week high 0.0044 -0.0180** (0.485) (0.042)

Target risks*Bidder 52-week high 0.0373*** (0.002)

Hostile -0.0104* -0.0103* -0.0106* (0.086) (0.088) (0.078)

Tender 0.0153*** 0.0151*** 0.0151*** (0.000) (0.000) (0.000)

Diversification 0.0010 0.0007 0.0012 (0.778) (0.855) (0.742)

Stock -0.0082* -0.0085* -0.0077 (0.095) (0.081) (0.114)

Cash 0.0298*** 0.0298*** 0.0296*** (0.000) (0.000) (0.000)

Relative size -0.0025 -0.0021 -0.0033 (0.599) (0.652) (0.492)

Bidder MV -0.0055*** -0.0051*** -0.0054*** (0.000) (0.000) (0.000)

Bidder M/B -0.0004 -0.0004 -0.0003 (0.319) (0.299) (0.371)

Bidder ROA 0.0435** 0.0464*** 0.0506*** (0.012) (0.007) (0.003)

Bidder leverage 0.0011 0.0012 0.0010 (0.448) (0.407) (0.502)

Constant 0.0238*** 0.0191** 0.0273*** (0.008) (0.025) (0.003)

N 2018 2018 2018

R2 0.092 0.092 0.098

Chapter 5. Tables

282

Table 5.6: Whether market-anticipated risks increase offer premiums?

This table presents results of OLS regressions of offer premiums on market anticipation driven risks. Offer

premiums are defined as the logarithmic difference between the offer price and the target price 30 days prior to

the takeover announcement date, which is the same as Baker et al. (2012). Bidder 52-week high is the bidder

reference point (BRP), target risks is the dummy variable, taking value of 1 if targets are above the median value

of their idiosyncratic volatility and below the median value of their stock price, 0 otherwise. The first regression

presents results of offer premiums on target risk, the second regression presents results of offer premiums on

bidder 52-week high, and the third regression presents results of offer premiums on market anticipation driven

risks, which is an interaction variable between target risk and bidder 52-week high. Control variables in this panel

are as the same as those in specification (5) of Table 5.4. Variable definitions are as in the notes of Table 5.2. P-

value is reported in parentheses. Statistical significance at the 1% level, 5% level, and 10% level, denoted ***, **

and * respectively.

Offer premiums (1) (2) (3)

Target risks 0.0615*** 0.0803*** (0.000) (0.000)

Bidder 52-week high 0.0544** 0.0958*** (0.016) (0.000)

Target risks*Bidder 52-week high -0.0856** (0.041)

Hostile 0.0379* 0.0398* 0.0383* (0.076) (0.064) (0.072)

Tender 0.0747*** 0.0727*** 0.0711*** (0.000) (0.000) (0.000)

Diversification -0.0227* -0.0227* -0.0228* (0.094) (0.096) (0.092)

Stock 0.0138 0.0110 0.0087 (0.435) (0.534) (0.622)

Cash -0.0265 -0.0296* -0.0259 (0.104) (0.069) (0.111)

Relative size 0.1499*** 0.1491*** 0.1477*** (0.000) (0.000) (0.000)

Bidder MV 0.0736*** 0.0752*** 0.0736*** (0.000) (0.000) (0.000)

Bidder M/B 0.0013 0.0012 0.0011 (0.346) (0.398) (0.415)

Bidder ROA -0.1001 -0.0892 -0.0951 (0.124) (0.170) (0.149)

Bidder leverage -0.0061 -0.0056 -0.0051 (0.268) (0.318) (0.354)

Target MV -0.0883*** -0.0967*** -0.0874*** (0.000) (0.000) (0.000)

Target M/B -0.0000 -0.0003 -0.0003 (0.986) (0.871) (0.862)

Target ROA 0.1388*** 0.1277*** 0.1397*** (0.001) (0.002) (0.001)

Target leverage -0.0072* -0.0077* -0.0071 (0.098) (0.080) (0.103)

Constant 0.1425*** 0.1834*** 0.1206*** (0.000) (0.000) (0.001)

N 1869 1869 1869

R2 0.163 0.162 0.169

Chapter 5. Tables

283

Table 5.7: Do market-anticipated risks push the firm to work hard?

This table presents results of OLS regression of bidder BHARs on target risk. BHARs are calculated with the

market-adjusted model. We include an interaction variable between target risk and bidder 52-week high in

specification (3). Variable definitions are as in the notes of Table 5.2. P-value is reported in parentheses. Statistical

significance at the 1% level, 5% level, and 10% level, denoted ***, ** and * respectively.

BHAR36m (1) (2) (3)

Target risks 0.1078** -0.0310 (0.025) (0.621)

Bidder 52-week high 0.1900*** -0.0513 (0.006) (0.491)

Target risks*Bidder 52-week high 0.3571*** (0.004)

Hostile 0.1262 0.1230 0.1200 (0.172) (0.180) (0.188)

Tender 0.0405 0.0229 0.0308 (0.387) (0.622) (0.507)

Diversification -0.0789* -0.0716* -0.0788** (0.050) (0.070) (0.050)

Stock -0.1631*** -0.1730*** -0.1655*** (0.003) (0.002) (0.002)

Cash -0.0016 0.0001 -0.0014 (0.973) (0.998) (0.977)

Relative size 0.0684 0.0446 0.0557 (0.162) (0.351) (0.262)

Bidder MV 0.0220** 0.0164* 0.0232** (0.034) (0.093) (0.023)

Bidder M/B -0.0118*** -0.0120*** -0.0117*** (0.000) (0.000) (0.000)

Bidder ROA 0.5056*** 0.5511*** 0.6084*** (0.001) (0.000) (0.000)

Bidder leverage 0.0139 0.0160 0.0160 (0.394) (0.331) (0.332)

Constant -0.2137** -0.1769* -0.2090** (0.030) (0.053) (0.034)

N 1867 1867 1867

R2 0.038 0.042 0.048

Chapter 5. Tables

284

Table 5.8: Robustness tests: Various reference candidates in relation to the firm level

This table presents results of logistic regressions of target risk on selected reference point candidates. Dependent

variable is a dummy variable, taking value of 1 if targets are above the median values of their idiosyncratic

volatility and below the median value of their stock price ending 30 days prior to the announcement date, 0

otherwise. Control variables are the same as those in specification (5) of Table 5.3. Bidder historical high refers

to the extent to which the current stock price deviated from the firm’s highest stock price, measured with

logarithmic term difference between the firm’s historical high and its current price. We collected all U.S. listed

firms whose stock price between 1958 and 2014 based on data availability. Bidder recency ratio is defined as the

days since the firm’s 52-week high arrives, measured with 1 minus the number of days since the firm’s 52-week

highest stock price arrives divided by 365 days. Prior market loss and prior ROA loss is dummy variables, taking

value of 1 if the firm’s market reaction (ROA) is lower than that of the past year, 0 otherwise. Specification (1)

presents results of target risks on bidder historical high. Specification (2) presents results of target risks on bidder

recency ratio. Specification (3) presents results of target risks on bidder prior market loss. Specification (4)

presents results of target risks on bidder ROA loss. Specification (5) presents results of target risks on bidder 52-

week high (BRP), controlling for those reference point candidates as well as those standard variables highlighted

in the M&A literature. Control variables in each regression are the same as those in specification (5) of Table 5.4.

Variable definitions are as in the notes of Table 5.2. P-value is reported in parentheses. Statistical significance at


Chapter 5. Tables

285


Logit model. Target risks (1) (2) (3) (4) (5)

Bidder historical high 0.434*** 0.197

(0.000) (0.115)

Bidder recency ratio -0.694*** -0.011

(0.002) (0.969)

Prior market loss 0.464*** 0.154

(0.003) (0.411)

Prior ROA loss 0.104 0.075

(0.468) (0.608)

Bidder 52-week high 0.997***

(0.000)

Hostile 0.263 0.225 0.194 0.169 0.218

(0.400) (0.464) (0.525) (0.591) (0.468)

Tender 0.015 0.081 0.092 0.103 0.002

(0.933) (0.655) (0.612) (0.572) (0.992)

Diversification 0.127 0.176 0.167 0.179 0.150

(0.439) (0.282) (0.309) (0.277) (0.364)

Stock 0.088 0.041 0.104 0.078 0.054

(0.669) (0.844) (0.614) (0.703) (0.795)

Cash -0.557*** -0.611*** -0.581*** -0.555*** -0.570***

(0.005) (0.003) (0.004) (0.005) (0.005)

RelativeSize 0.163 0.183 0.195 0.189 0.155

(0.418) (0.349) (0.324) (0.336) (0.448)

Bidder MV 0.105* 0.111* 0.110** 0.092 0.115**

(0.063) (0.050) (0.049) (0.100) (0.047)

Bidder M/B -0.003 -0.004 -0.002 -0.008 -0.003

(0.824) (0.710) (0.864) (0.466) (0.781)

Bidder ROA 0.255 -0.228 -0.194 -0.079 0.154

(0.646) (0.679) (0.725) (0.888) (0.794)

Bidder Leverage -0.014 -0.020 -0.031 -0.024 -0.010

(0.851) (0.794) (0.676) (0.746) (0.898)

Target MV -1.495*** -1.506*** -1.507*** -1.490*** -1.509***

(0.000) (0.000) (0.000) (0.000) (0.000)

Target M/B -0.040** -0.042** -0.045** -0.046** -0.039**

(0.034) (0.038) (0.027) (0.020) (0.044)

Target ROA -2.267*** -2.400*** -2.408*** -2.486*** -2.210***

(0.000) (0.000) (0.000) (0.000) (0.000)

Target Leverage 0.048 0.048 0.055 0.056 0.044

(0.334) (0.347) (0.291) (0.275) (0.378)

Year Y Y Y Y Y

Industry Y Y Y Y Y

Constant 5.697*** 6.431*** 5.842*** 6.086*** 5.467***

(0.000) (0.000) (0.000) (0.000) (0.000)

N 2018 2018 2018 2018 2018

pseudo R2 0.512 0.510 0.509 0.506 0.520

Chapter 5. Tables

286

Table 5.9: Robustness test: Do the market returns or peer pressures play a role in the firm’s risk-taking?

This table presents results of logistic regressions of target risk on other variables that have influence on risk

preferences. Dependent variable is a dummy variable, taking value of 1 if targets are above the median values of

their idiosyncratic volatility and below the median value of their stock price ending 30 days prior to the

announcement date, 0 otherwise. Market 52-week high is defined as logarithmic term difference between the

market index 52-week high and the market index 30 days prior to the takeover announcement date. Peer pressure

reflects the extent to which the individual firm’s 52-week high deviated from the industry 52-week high median.

It is a dummy variable, taking value of 1 if the firm’s 52-week high lower than the industry’s 52-week high median,

0 otherwise. Specification (1) report results of target risks on the market 52-week high. Specification (2) reports

results of target risks on the firm’s peer pressure. Specification (3) report target risks on the bidder reference point

which is the bidder 52-week high, controlling for both the market 52-week high and the industry’s pressure. All

specifications control for a series of standard variables highlighted in the prior M&A literature, which are the same

as those presented in specification (5) of Table 5.4. Variable definitions are as in the notes of Table 5.2. P-value is

reported in parentheses. Statistical significance at the 1% level, 5% level, and 10% level, denoted ***, ** and *

respectively.

Chapter 5. Tables

287

Table 5.9. (Continued)

Logit model. Target risks (1) (2) (3)

Market 52-week high 2.060* 0.329

(0.086) (0.786)

Peer pressure 0.622*** 0.079

(0.000) (0.700)

Bidder 52-week high 1.167***

(0.000)

Hostile 0.201 0.188 0.188

(0.522) (0.534) (0.529)

Tender 0.107 0.085 0.026

(0.553) (0.638) (0.886)

Diversification 0.191 0.185 0.178

(0.244) (0.264) (0.283)

Stock 0.077 0.022 0.026

(0.706) (0.914) (0.901)

Cash -0.566*** -0.581*** -0.559***

(0.004) (0.004) (0.005)

RelativeSize 0.181 0.199 0.153

(0.352) (0.314) (0.452)

Bidder MV 0.088 0.118** 0.108*

(0.113) (0.040) (0.062)

Bidder M/B -0.007 -0.007 -0.007

(0.520) (0.519) (0.535)

Bidder ROA -0.154 -0.201 -0.051

(0.781) (0.721) (0.931)

Bidder Leverage -0.026 -0.009 -0.006

(0.724) (0.901) (0.935)

Target MV -1.490*** -1.496*** -1.501***

(0.000) (0.000) (0.000)

Target M/B -0.045** -0.045** -0.043**

(0.027) (0.022) (0.032)

Target ROA -2.478*** -2.320*** -2.283***

(0.000) (0.000) (0.000)

Target Leverage 0.052 0.047 0.045

(0.310) (0.342) (0.364)

Year Y Y Y

Industry Y Y Y

Constant 6.136*** 5.651*** 5.669***

(0.000) (0.000) (0.000)

N 2018 2018 2018

pseudo R2 0.507 0.512 0.518

Chapter 5. Tables

288

Table 5.10: Robustness test: Subsample tests

This table presents results of logistic regression of target riskiness on bidder 52-week high by different deal

samples. We partition the whole sample by the method of payment, whether the deal is a tender offer, whether the

deal is diversified, and whether the deal is successful. Control variables are as the same as those in specification

(5) of Table 5.4, and suppressed for the sake of brevity. Variable definitions are as in the notes of Table 5.2. P-

value is reported in parentheses. Statistical significance at the 1% level, 5% level, and 10% level, denoted ***, **

and * respectively.

Logit model. Target risks Payment methods Tender offer

Cash Stock Mix Yes No

Bidder 52-week high 2.012*** 0.467 2.118*** 2.470*** 0.909*** (0.000) (0.279) (0.000) (0.000) (0.003)

Controls Y Y Y Y Y

Constant 5.981*** 8.711*** 3.877 7.185*** 5.985*** (0.000) (0.000) (0.162) (0.000) (0.000)

N 718 629 556 509 1482

Pseudo R2 0.540 0.544 0.585 0.561 0.534 Diversification Successful

Yes No Yes No

Bidder 52-week high 0.620 1.641*** 1.129*** 2.312***

(0.247) (0.000) (0.000) (0.002)

Controls Y Y Y Y

Constant 5.474*** 6.285*** 6.063*** 7.688***

(0.000) (0.000) (0.000) (0.000)

N 659 1357 1587 395

Pseudo R2 0.520 0.545 0.524 0.606

Chapter 5. Tables

289

Table 5.11: Robustness test: OLS regressions of target idiosyncratic risk on the bidder reference point

This table presents results of OLS regression of target idiosyncratic risks (TIR) on the bidder 52-week high (BRP),

where target idiosyncratic risks are obtained from different estimation model used in the literature. The dependent

variable of the first specification is total volatility of the target, which is the standard deviation of the target’s

stock returns in the period of 1 month and 6 months prior to the takeover announcement date. Target idiosyncratic

risks of specification (2) are calculated with the residuals obtained by the market model (MM_TIR). Target

idiosyncratic risks of specification (3) are calculated with the residuals from Fama-French 4 factor model

(FF4factor_TIR). Variable definitions are as in the notes of Table 5.2. All regressions include the year and industry

dummies. P-value is reported in parentheses. Statistical significance at the 1% level, 5% level, and 10% level,

denoted ***, ** and * respectively.

OLS regressions (1) (2) (3)

Total Volatility MM_TIR FF4factor_TIR

BRP 0.009*** 0.010*** 0.010***

(0.000) (0.000) (0.000)

Hostile -0.002* -0.003** -0.003**

(0.086) (0.017) (0.020)

Tender -0.001 -0.001 -0.001

(0.150) (0.121) (0.126)

Diversification 0.000 0.000 0.000

(0.667) (0.674) (0.657)

Stock 0.003*** 0.002** 0.002**

(0.003) (0.050) (0.049)

Cash -0.002* -0.002** -0.003***

(0.059) (0.011) (0.010)

RelativeSize 0.001 0.000 0.001

(0.557) (0.578) (0.564)

Bidder MV 0.001*** 0.001** 0.001**

(0.007) (0.026) (0.026)

Bidder M/B 0.000*** 0.000*** 0.000***

(0.002) (0.001) (0.001)

Bidder ROA -0.013*** -0.010*** -0.010***

(0.000) (0.008) (0.008)

Bidder Leverage 0.000 -0.000 -0.000

(0.988) (0.474) (0.463)

Target MV -0.005*** -0.006*** -0.006***

(0.000) (0.000) (0.000)

Target M/B 0.000*** 0.000* 0.000*

(0.009) (0.079) (0.087)

Target ROA -0.022*** -0.024*** -0.024***

(0.000) (0.000) (0.000)

Target Leverage -0.000 -0.000 -0.000

(0.980) (0.789) (0.812)

Year Y Y Y

Industry Y Y Y

Constant 0.041*** 0.046*** 0.046***

(0.000) (0.000) (0.000)

N 2018 2018 2018

R2 0.610 0.599 0.599

Chapter 5. Tables

290

Table 5.12: Robustness test: OLS regressions of bidder CARs on market-anticipated risks

This table reports the results of OLS regressions of bidder announcement returns on the market’s anticipated risks.

We measure the market’s anticipated risks with an interactive term between target risks and the bidder reference

point (BRP). The first three specifications measure the relation between bidder cumulative abnormal returns on 5

days around the announcement date and calculated with the market model (CAR5mm) while the last three

specifications reflect the relation between bidder cumulative abnormal returns on 3 days around the announcement

date and calculated with the market-adjusted model (CAR3ma). Variable definitions are as in the notes of Table

5.2. P-value is reported in parentheses. Statistical significance at the 1% level, 5% level, and 10% level, denoted

***, ** and * respectively.

OLS regressions (1) (2) (3) (4) (5) (6)

Bidder CAR5mm Bidder CAR3ma

Target risks 0.0015 -0.0134** -0.0028 -0.0111**

(0.740) (0.017) (0.485) (0.026)

BRP 0.0106 -0.0155 -0.0013 -0.0177**

(0.150) (0.115) (0.837) (0.041)

Target risks*BRP 0.0418*** 0.0275**

(0.003) (0.021)

Hostile -0.0120* -0.0120* -0.0123* -0.0109* -0.0109* -0.0111*

(0.080) (0.083) (0.076) (0.070) (0.071) (0.065)

Tender 0.0190*** 0.0184*** 0.0186*** 0.0140*** 0.0142*** 0.0142***

(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)

Diversification -0.0003 -0.0004 -0.0002 0.0007 0.0004 0.0008

(0.934) (0.915) (0.968) (0.854) (0.912) (0.824)

Stock -0.0103* -0.0110** -0.0100* -0.0078 -0.0078 -0.0071

(0.062) (0.045) (0.065) (0.108) (0.108) (0.138)

Cash 0.0259*** 0.0259*** 0.0256*** 0.0290*** 0.0291*** 0.0289***

(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)

RelativeSize -0.0069 -0.0073 -0.0079 -0.0032 -0.0026 -0.0035

(0.187) (0.153) (0.127) (0.513) (0.581) (0.468)

Bidder MV -0.0062*** -0.0060*** -0.0061*** -0.0054*** -0.0052*** -0.0054***

(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)

Bidder M/B -0.0008* -0.0008* -0.0008* -0.0001 -0.0001 -0.0000

(0.074) (0.066) (0.088) (0.828) (0.823) (0.901)

Bidder ROA 0.0623*** 0.0668*** 0.0721*** 0.0375** 0.0378** 0.0409**

(0.002) (0.001) (0.000) (0.026) (0.024) (0.015)

Bidder Leverage 0.0023 0.0024 0.0022 0.0007 0.0007 0.0005

(0.179) (0.154) (0.193) (0.655) (0.642) (0.728)

Constant 0.0305*** 0.0272*** 0.0332*** 0.0257*** 0.0232*** 0.0294***

(0.002) (0.005) (0.001) (0.004) (0.006) (0.001)

N 2017 2017 2017 2018 2018 2018

R2 0.091 0.092 0.098 0.083 0.083 0.086

Chapter 5. Tables

291

Table 5.13: Robustness test: OLS regressions of BHARs on market-anticipated risks

This table reports the results of OLS regressions of bidder buy-and-hold abnormal returns (BHARs) on the

market’s anticipated risks. BHARs are calculated with the market-adjusted model. Dependent variable of the first

three specifications is BHARs over 12 months after the completion of a merger (BHAR12ma), while dependent

variable of the last three specifications is BHARs over 24 months after the completion of a merger (BHAR24ma).

Variable definitions are as in the notes of Table 5.2. P-value is reported in parentheses. Statistical significance at


(1) (2) (3) (4) (5) (6)

BHAR12ma BHAR24ma

Target risks 0.0387 -0.0113 0.0489 -0.0143

(0.149) (0.742) (0.196) (0.767)

BRP 0.0244 -0.0820* 0.0522 -0.0729

(0.525) (0.071) (0.273) (0.224)

Target risks*BRP 0.1565** 0.1843**

(0.032) (0.045)

Hostile -0.0301 -0.0309 -0.0322 0.0571 0.0560 0.0544

(0.497) (0.486) (0.469) (0.348) (0.359) (0.372)

Tender 0.0137 0.0102 0.0138 0.0241 0.0183 0.0225

(0.597) (0.692) (0.592) (0.515) (0.620) (0.543)

Diversification -0.0420* -0.0383* -0.0417* -0.0466 -0.0425 -0.0464

(0.064) (0.088) (0.066) (0.143) (0.177) (0.145)

Stock -0.0404 -0.0411 -0.0378 -0.0931** -0.0954** -0.0915**

(0.188) (0.183) (0.220) (0.024) (0.021) (0.028)

Cash 0.0085 0.0087 0.0080 0.0077 0.0081 0.0073

(0.744) (0.739) (0.758) (0.835) (0.826) (0.842)

RelativeSize 0.0181 0.0099 0.0152 0.0393 0.0287 0.0348

(0.531) (0.724) (0.602) (0.309) (0.441) (0.372)

Bidder MV 0.0091 0.0060 0.0092 0.0195** 0.0161** 0.0198**

(0.114) (0.278) (0.111) (0.015) (0.037) (0.014)

Bidder M/B -0.0039* -0.0039* -0.0037* -0.0122*** -0.0122*** -0.0120***

(0.055) (0.056) (0.065) (0.000) (0.000) (0.000)

Bidder ROA 0.2911*** 0.2888*** 0.3145*** 0.4589*** 0.4651*** 0.4951***

(0.004) (0.003) (0.001) (0.000) (0.000) (0.000)

Bidder Leverage -0.0022 -0.0023 -0.0023 0.0122 0.0125 0.0125

(0.830) (0.821) (0.821) (0.326) (0.321) (0.320)

Constant -0.0886 -0.0561 -0.0724 -0.1835** -0.1517** -0.1700**

(0.126) (0.305) (0.220) (0.021) (0.042) (0.035)

N 1867 1867 1867 1867 1867 1867

R2 0.018 0.018 0.022 0.040 0.040 0.043

292

CHAPTER 6: CONCLUSIONS

Chapter 6. Conclusions

293

6.1. Main findings

This thesis has investigated the reference point effect on M&As. The thesis allows the

researcher to assess investors’ major investment decision-making processes in condition of

risks. The thesis uses the firms’ 52-week highs as reference points since they are frequently

reported by the media that shape investors’ minds, used by M&A participants to assess the

M&A motive and performance. This thesis has contributed to M&A literature by enhancing

this new strand of behavioural finance theory of M&As. It has become evident that firms’ 52-

week highs are as important indications for the firm’s bargaining power, managerial market-

timing, and risk-taking.

The reference point theory has been investigated into many key aspects of M&As in this thesis.

First, the target 52-week high effect was first applied in the U.K. setting, it was found that the

reference point effect reinforces the target firm’s bargaining power in the scenarios where the

takeover market is more competitive and bidders endure extensive information barriers.

Secondly, both the bidder and the target reference prices were revealed to give important

indications of the firm’s valuation to the market and the manager alike. The reference point

theory accommodates implications of Shleifer and Vishny’s misvaluation hypothesis (2003),

suggesting that bidder managers can time the market with the reference points. Thirdly, direct

evidence was produced that the bidder reference price reflects the market risk appetiser, which

is transferrable when bidder managers make M&A decisions according to market-anticipated

risks. Besides, it was established that the reference point theory has strong explanatory value

on many key aspects of M&As.

Chapter 3 has investigated a sample of 451 domestic and 155 cross-border public M&As in the

United Kingdom, affirming that the target reference point plays an essential role in bargaining


294

power. The results obtained in this chapter are consistent with Baker et al.’s U.S. findings

(2012). It is suggested that the U.K. target is much easier to reinforce the bargaining power

with its reference price, based on the rationale that the market is competitive compared with

the United States. Unlike U.S. firms, whose corporate governance system is manager-oriented,

U.K. firms have a shareholders-oriented corporate governance system where the reference

point effect is stronger since managerial decisions are likely to rely on the reference point to

convince their shareholders that interests of the two sides are closely aligned.

According to these predictions, Chapter 3 has found a positive relationship between the target

reference price and offer premiums, suggesting that the reference point effect is translated into

a strong bargaining power of the target for high M&A offer premiums. Further analyses of the

market reactions to bidders whose payment according to the target reference point show

evidence of overpayment among domestic bidders, whereas there is little evidence of

overpayment among cross-border bidders, suggesting that the market takes foreign bidders’

M&A offer premiums according to the target reference point for granted, unlike those paid by

domestic bidders. Two possible explanations were advisable for this finding. First, domestic

bidders are believed to have a more sophisticated networking than foreign bidders, meaning

that they should have more private information to assess the value of the target firm to reduce

M&A offer premiums, but paying according to the target reference price gives the market an

indication that domestic bidders maybe overconfident. Secondly, it is believed that cross-

border bidders in order to enter a more competitive market to protect their shareholders’ rights

tend to have a relatively weak bargaining position, thus offering a price in favour of the target.

Chapter 3 has made two distinct contributions to the reference point theory of M&As. First,

scenarios where the reference point effect was enhanced were found. The reference point effect


295

is strong in a more competitive market, as compared the findings of this chapter with those of

the United States. The reference point is pronounced in the case of bidders lacking private

information about the target. Secondly, the reference point theory is used to distinguish the

M&A motives between cross-border and domestic bidders. The market regards domestic

bidders’ offer premiums according to the target reference point as an overpayment, whilst

supporting that of cross-border bidders whose offer premiums rely on the target reference point.

Chapter 4 has thoroughly explored the reference point theory in M&A misvaluation. The

reference point prices of the two firms involved were taken into considerations. A proxy of the

relative reference point (RRP) was constructed to explain how bidders time the market. The

results indicate that stocks are 1.04% more likely to be used for every 10% increase in the RRP,

suggesting that bidders use stocks to dilute overvaluation when there is a sign of it according

to the RRP. The findings also show that for a 10% increase in the RRP lead to an increase of

M&A offer premiums by approximately 0.9%, suggesting that the sign of relative valuation

make takeovers costly. Despite paying according to the RRP incurring negative announcement

returns, offer premiums are positively related to the long-run performance of the stock bidders,

indicating the RRP facilities market-timing.

Chapter 4 has made three contributions to M&A literature. First, it reconciles the misvaluation

hypothesis with the reference point theory of M&As. A new proxy was proposed, based on the

market perception of a firm’s valuation, to explain M&A misvaluation. The idea of using the

RRP for misvaluation eliminates any concerns arising from the biases caused by a firm’s

fundamentals. Secondly, direct evidence is offered showing that not only the market but also

the managers are subject to reference-dependence bias. But they were found to treat the

reference point differently, as the market reacts negatively to bid announcements holding that


296

bidders who pay according to the RRP are paying too much, whilst managers are concerned

themselves with the interests of the shareholders in the long run. By looking at the RRP, bidders

recognise the relatively more overvalued stocks and in exchange for less negative long-run

performance. Thirdly, the RRP justifies the rationale of payment method choices. Lower M&A

offer premiums are associated with cases where bidders pay with stocks instead of cash for

larger RRP acquisitions, and conversely bidders paying with cash for a lower RRP acquisition

than for a higher RRP acquisition.

Chapter 5 has examined the role of the bidder reference point in the managerial risk-taking

behaviour. One of the most important implications of the reference point is that people have a

strong tendency to avoid loss. This implication was tested in the M&A context and the bidder

reference point was proposed, which represents the extent to which a bidder’s current price

deviates from the bidder 52-week high, capturing the market anticipation for a firm’s

investment decisions. A significantly decline in performance relative to the bidder 52-week

high motivates a great risk appetizer for the market, which also motivates the firm to take on

risky projects so as to recover the loss. The measurement of target risks follows the proxy of

Kumar (2009). It became clear that bidder managers select targets according to the market-

anticipated risks and are rewarded by the market.

Specifically, Chapter 5 has found a positive relationship between the riskiness of an M&A

decision and the bidder reference point, suggesting that the market anticipates the M&A

decision. A positive relationship between bidder announcement returns and the market-

anticipated risks was emerged, showing that bidders listening to the market would justify their

M&A motives. It is also believed that bidders taking anticipated risks expose themselves to

shareholders’ monitoring, pushing them to exhibit sufficient effort observed by the market. The


297

finding was verified by studying the bidder reference point effect on M&A offer premiums and

the bidder long-term performance. Offer premiums were found to be significantly lower when

M&A decisions are based on the market-anticipated risks, suggesting that managers who link

their interests with those of their shareholders exhibit great efforts in negotiation. Further, the

market-anticipated risks are positively related with long-term performance, suggesting that

managers in the face of loss have a strong incentive to prove themselves good managers to

avoid the risk of being replaced and so they work harder to integrate the firm’s resources.

Chapter 5 has made two distinct contributions to previous literature. First, it uses the reference

point theory to address managerial risk-taking behaviour in the M&A context. Managers take

risks according to the bidder reference point, suggesting that the stock market serves as a risk

transfer mechanism. Secondly, the market-anticipated risks were distinguished from

unanticipated risks (i.e. risks driven by managerial overconfidence and agency problems),

resulting in negative market reactions. Managers undertaking M&As are able to justify their

motives by following the market-anticipated risks, and receive positive market reactions

accordingly. It is clear that decisions made upon the market-anticipated risks improve a firm’s

M&A performance.

On the whole, this thesis has enhanced Baker et al.’s findings (2012) and explored reference

point theory of M&A in a number of ways. The principal findings of this thesis suggest that

the target 52-week high represents a role of the target’s bargaining power. When applying the

reference point theory to the misvaluation hypothesis of M&As, the joint effects of the target

and bidder reference point allow an assessment of the way of managers timing the market.

Despite taking too many unanticipated risks in M&As end up value-destroying to the firm

according to eariler M&A literature, Chapter 5 indicates that managers can rationalise their


298

risk-taking incentive by focusing on the bidder reference point, and performing this practise

with caution and efforts can improve the firm’s M&A performance.

The main findings of this thesis show that both institutional and individual investors are subject

to reference-dependence bias, suggesting that it is a natural human nature response to make

risky decisions in accordance with the reference point. However, managers and the market may

value a firm differently even with the same information, as one of the main findings suggested

in Chapter 4. Findings in Chapter 3 suggest that M&As are not only a bargaining game between

the management teams of the two firms, but also a trade-off game between the managers and

the shareholders in a firm. It is proposed that managers should always give a thought for the

market when making major investment decisions. As Chapter 5 suggested, M&A decisions

made based on the market’s anticipation convince shareholders that the firm works hard to

protect the value of the firm, suggesting that managers would do whatever the market thinks

right in the face of loss in firm’s performance.

6.2. Implications, limitations and recommendations for future research

Although the research was fruitful in finding results regarding the reference point effect on

M&As, it is subject to a number of limitations, implying some proposals for future research.

First, due to data availability, Chapter 3 has a small fraction of cross-border acquisitions in the

U.K. market. Therefore, the reference point effect should be explored further by extending the

sample into a worldwide cross-border acquisition sample. Further research should take into

account some macro-economic factors, such as total trading volumes, and stock markets of the

two countries whose firms are involved in the cross-border acquisition. Firms in a country with

relatively strong in trading and high in stock market quality are more easily to reinforce the

bargaining power, which is in line with Erel et al. (2012). It remains unexplored whether such


299

macro-economic factors make bidders to be less likely to focus on the target 52-week high,

given the possibility that target shareholders are also willing to sell the firm to a relatively

strong market regardless of the reference point effect. Meanwhile, the fact that the reference

point effect is enhanced in a setting where the corporate governance system is strong is taken

for granted in this thesis. However, it should be noted that firms’ corporate governance quality

varies significantly, which requires inclusion of some corporate governance related proxies in

the firm level.

Secondly, Chapter 5 has employed a number of reference point candidates highlighted in

previous literature as robustness tests for the bidder 52-week high. The regressions gave a

simple indication that the firm 52-week high is the most suitable reference point effect. It

should be expected that managers consider various reference point candidates, which requires

the researcher to construct a composite index of reference points, including the firm-, the

market-, and the sector-levels. By doing so, one should assign a weight for relevant reference

point candidate, similar to an approach used by Hayward and Hambrick (1997), who

constructed a composite index for the hubris management. In addition, this thesis only looks at

public acquisitions since only publicly-listed firms have available 52-week high. However, it

is worth noting that private acquisitions in which bidders encounter higher levels of information

asymmetry should be a more suitable testing ground for the reference point effect. Although it

is hard to find the reference point in private targets, it is proposed that efforts should be made

to explore the reference point price of the bidder, for example, the unique characteristics of the

bidders have undertaken many private acquisitions in the past.

Finally, Chapter 5 has focused on the risk-taking incentive from the perspective of the market.

The risk tolerance of the manager should also be taken into consideration. The consequence is


300

the question whether the firm that made a failed deal in the past is likely to take on a risky deal,

as opposed to one that have made a successful deal. In other words, the completion status of

the prior deal may be employed as a reference point for a current M&A decision. Though

Chapter 5 suggests that market-anticipated risks improve a firm’s M&A performance, it

remains unexplored that the market-unanticipated risks effect on the firm’s M&A performance

by including managerial overconfidence and corporate governance related proxies. This

Chapter also draws several implications for practitioners. Managers should be concerned about

stock market signals while making major investment decisions, since the market will inform

the firm how many risks it is willing the firm to take. Managers who do not listen to the market

will be judged as overconfidence or irresponsibility. The researcher shows that taking risks

does not necessarily lead to bad outcomes if that they are justifiable. Though quantifying risk

magnitude is not an easy task according to March and Shapira’s view (1987), this researcher

indicates a possible timing for managerial investment decisions, which is largely explained in

the framework of the reference point theory of M&As.

Overall, the idea of using a firm’s reference points should also be applied in certain areas where

managers have an opportunity to time the market, such as SEOs, equity issuance, and in the

settings where the levels of information asymmetry are high.

301

REFERENCES

References

302

References

Agrawal, A., Jaffe, J.F., 2000. The post-merger performance puzzle: Advances in mergers and acquisitions.

Emerald Group Publishing Limited, pp. 7-41.

Agrawal, A., Jaffe, J.F., 2003. Do takeover targets underperform? Evidence from operating and stock returns.

Journal of Financial and Quantitative Analysis 38, 721-746.

Agrawal, A., Jaffe, J.F., Mandelker, G.N., 1992. The post‐merger performance of acquiring firms: A re‐

examination of an Anomaly. Journal of Finance 47, 1605-1621.

Aktas, N., De Bodt, E., Roll, R., 2009. Learning, hubris and corporate serial acquisitions. Journal of Corporate

Finance 15, 543-561.

Alexandridis, G., Fuller, K.P., Terhaar, L., Travlos, N.G., 2013. Deal size, acquisition premia and shareholder

gains. Journal of Corporate Finance 20, 1-13.

Alexandridis, G., Mavrovitis, C. F., & Travlos, N. G., 2012. How have M&As changed? Evidence from the sixth

merger wave. The European Journal of Finance, 18, 663-688.

Alexandridis, G., Petmezas, D., Travlos, N.G., 2010. Gains from mergers and acquisitions around the world: New

evidence. Financial Management 39, 1671-1695.

Alshwer, A.A., Sibilkov, V., Zaiats, N.S., 2011. Financial constraints and the method of payment in mergers and

acquisitions.

Andrade, G., Mitchell, M., Stafford, E., 2001. New evidence and perspectives on mergers.

Andreou, P.C., Louca, C., Petrou, A.P., 2016. Organizational learning and corporate diversification performance.

Journal of Business Research 69, 3270-3284.

Ang, J.S., Cheng, Y., 2006. Direct evidence on the market‐driven acquisition theory. Journal of Financial

Research 29, 199-216.

Antoniou, A., Arbour, P., Zhao, H., 2008. How much is too much: are merger premiums too high? European

Financial Management 14, 268-287.

Antoniou, A., Petmezas, D., Zhao, H., 2007. Bidder gains and losses of firms involved in many acquisitions.

Journal of Business Finance & Accounting 34, 1221-1244.

Armour, J., Skeel, D.A., 2007. The divergence of US and U.K. takeover regulation. Regulation 30, 08-24.

References

303

Asquith, P., 1983. Merger bids, uncertainty, and stockholder returns. Journal of Financial Economics 11, 51-83.

Asquith, P., Bruner, R.F., Mullins, D.W., 1983. The gains to bidding firms from merger. Journal of Financial

Economics 11, 121-139.

Avery, M., 2013. Literature Review for Local Polynomial Regression. Unpublished manuscript, Department of

Statistics, North Carolina State University, Raleigh, NC.

Baker, M., Pan, X., Wurgler, J., 2012. The effect of reference point prices on mergers and acquisitions. Journal

of Financial Economics 106, 49-71.

Baker, M., Wurgler, J., 2012. Behavioral corporate finance: A current survey. Handbook of the Economics of

Finance, 2, edited by George M. Constantinides, Milton Harris, and Rene M. Stulz. Handbooks in Economics.

New York, 2012.

Barber, B.M., Lyon, J.D., 1997. Detecting long-run abnormal stock returns: The empirical power and specification

of test statistics. Journal of Financial Economics 43, 341-372.

Barber, B.M., Odean, T., 2008. All that glitters: The effect of attention and news on the buying behavior of

individual and institutional investors. Review of Financial Studies 21, 785-818.

Barberis, N., Shleifer, A., Vishny, R., 1998. A model of investor sentiment. Journal of Financial Economics 49,

307-343.

Barberis, N., Xiong, W., 2009. What drives the disposition effect? An analysis of a long-standing preference-

based explanation. Journal of Finance 64, 751-784.

Barberis, N.C., 2013. Thirty years of prospect theory in economics: A review and assessment. Journal of

Economic Perspectives 27, 173-195.

Bargeron, L., Schlingemann, F.P., Stulz, R.M., Zutter, C.J., 2010. Are acquisition premiums lower because of

target CEOs’ conflicts of interest? SSRN Working Paper.

Bena, J., Li, K., 2014. Corporate innovations and mergers and acquisitions. Journal of Finance 69, 1923-1960.

Benartzi, S., Thaler, R.H., 1995. Myopic loss aversion and the equity premium puzzle. Quarterly Journal of


Ben-David, I., Drake, M.S., Roulstone, D.T., 2015. Acquirer valuation and acquisition decisions: Identifying

mispricing using short interest. Journal of Financial and Quantitative Analysis 50, 1-32.

References

304

Berger, P.G., Ofek, E., 1995. Diversification’s effect on firm value. Journal of Financial Economics 37, 39-65.

Berkovitch, E., Khanna, N., 1991. A theory of acquisition markets: Mergers versus tender offers, and golden

parachutes. Review of Financial Studies 4, 149-174.

Berkovitch, E., Narayanan, M., 1993. Motives for takeovers: An empirical investigation. Journal of Financial

and Quantitative Analysis 28, 347-362.

Bertrand, M., Mullainathan, S., 2003. Enjoying the quiet life? Corporate governance and managerial preferences.

Journal of Political Economy 111, 1043-1075.

Betton, S., Eckbo, B.E., Thorburn, K.S., 2008. Markup pricing revisited. Unpublished working paper, Tuck

School of Business at Dartmouth, Hanover, NH (2008-45).

Betton, S., Eckbo, B.E., Thorburn, K.S., 2009. Merger negotiations and the toehold puzzle. Journal of Financial


Bhootra, A., Hur, J., 2013. The timing of 52-week high price and momentum. Journal of Banking & Finance 37,

3773-3782.

Billett, M.T., Qian, Y., 2008. Are overconfident CEOs born or made? Evidence of self-attribution bias from

frequent acquirers. Management Science 54, 1037-1051.

Boone, A.L., Mulherin, J.H., 2007. How are firms sold? Journal of Finance 62, 847-875.

Borna, S., Lowry, J., 1987. Gambling and speculation. Journal of Business Ethics 6, 219-224.

Bouwman, C., Fuller, K., Nain, A., 2003. The performance of stock-price driven acquisitions. Social Science

Research Network. SSRN Working Paper.

Bouwman, C.H., Fuller, K., Nain, A.S., 2009. Market valuation and acquisition quality: Empirical evidence.

Review of Financial Studies 22, 633-679.

Bradley, M., Desai, A., Kim, E.H., 1983. The rationale behind interfirm tender offers: Information or synergy?

Journal of Financial Economics 11, 183-206.

Bradley, M., Desai, A., Kim, E.H., 1988. Synergistic gains from corporate acquisitions and their division between

the stockholders of target and acquiring firms. Journal of Financial Economics 21, 3-40.

References

305

Bradley, M., Sundaram, A., 2004. Do acquisitions drive performance or does performance drive acquisitions.

Social Science Research Network.

Brandenburger, A., Polak, B., 1996. When managers cover their posteriors: Making the decisions the market

wants to see. The RAND Journal of Economics 27, 523-541.

Brar, G., Giamouridis, D., Liodakis, M., 2009. Predicting European takeover targets. European Financial

Management 15, 430-450.

Brenner, R., 1983. History-The human gamble. University of Chicago Press: 1st Edition.

Brown, S.J., Warner, J.B., 1985. Using daily stock returns: The case of event studies. Journal of Financial

Economics 14, 3-31.

Burch, T.R., Nanda, V., Silveri, S., 2012. Do institutions prefer high-value acquires? An analysis of trading in

stock-financed acquisitions. Journal of Financial Research 35, 211-241.

Burghof, H.-P., Prothmann, F., 2011. The 52-week high strategy and information uncertainty. Financial Markets

and Portfolio Management 25, 345-378.

Chang, S., 1998. Takeovers of privately held targets, methods of payment, and bidder returns. Journal of Finance

53, 773-784.

Chikh, S., Filbien, J.-Y., 2011. Acquisitions and CEO power: Evidence from French networks. Journal of

Corporate Finance 17, 1221-1236.

Chira, I., Madura, J., 2015. Reference point theory and pursuit of deals. Financial Review 50, 275-300.

Clark, K., Ofek, E., 1994. Mergers as a means of restructuring distressed firms: An empirical investigation.


Comment, R., Jarrell, G.A., 1995. Corporate focus and stock returns. Journal of Financial Economics 37, 67-87.

Core, J., Guay, W., 2002. Estimating the value of employee stock option portfolios and their sensitivities to price

and volatility. Journal of Accounting Research 40, 613-630.

Cornu, P., 2000. The deterring role of the medium of payment in takeover contests: Theory and evidence from

the U.K.. European Financial Management 6, 423-440.

References

306

Croci, E., Petmezas, D., 2015. Do risk-taking incentives induce CEOs to invest? Evidence from acquisitions.

Journal of Corporate Finance 32, 1-23.

Croci, E., Petmezas, D., Travlos, N., 2012. Asymmetric information and target firm returns. European Journal of

Finance 18, 639-661.

Croci, E., Petmezas, D., Vagenas-Nanos, E., 2010. Managerial overconfidence in high and low valuation markets

and gains to acquisitions. International Review of Financial Analysis 19, 368-378.

Danbolt, J., 1995. An analysis of gains and losses to shareholders of foreign bidding companies engaged in cross-

border acquisitions into the United Kingdom, 1986–1991. European Journal of Finance 1, 279-309.

Danbolt, J., 2004. Target company Cross-border effects in acquisitions into the U.K.. European Financial

Management 10, 83-108.

Datta, D.K., Pinches, G.E., Narayanan, V., 1992. Factors influencing wealth creation from mergers and

acquisitions: A meta‐analysis. Strategic Management Journal 13, 67-84.

Datta, S., Iskandar-Datta, M., Raman, K., 2001. Executive compensation and corporate acquisition decisions.

Journal of Finance 56, 2299-2336.

De, S., Jindra, J., 2012. Why newly listed firms become acquisition targets. Journal of Banking & Finance 36,

2616-2631.

DeBondt, W.F., Thaler, R., 1985. Does the stock market overreact? Journal of Finance 40, 793-805.

Denis, D.J., Denis, D.K., Yost, K., 2002. Global diversification, industrial diversification, and firm value.


Devos, E., Kadapakkam, P.R., Krishnamurthy, S., 2009. How do mergers create value? A comparison of taxes,

market power, and efficiency improvements as explanations for synergies. Review of Financial Studies 22, 1179-

1211.

Di Giuli, A., 2013. The effect of stock misvaluation and investment opportunities on the method of payment in

mergers. Journal of Corporate Finance 21, 196-215.

Díaz, K.B.D., Azofra, S.S., Gutiérrez, C.L., 2009. Are M&A premiums too high? Analysis of a quadratic

relationship between premiums and returns. Quarterly Journal of Finance and Accounting 48, 5-21.

References

307

Dickerson, A.P., Gibson, H.D., Tsakalotos, E., 2003. Is attack the best form of defence? A competing risks

analysis of acquisition activity in the U.K.. Cambridge Journal of Economics 27, 337-357.

Dodd, P., 1980. Merger proposals, management discretion and stockholder wealth. Journal of Financial


Dodd, P., Ruback, R., 1977. Tender offers and stockholder returns: An empirical analysis. Journal of Financial


Dong, M., Hirshleifer, D., Richardson, S., Teoh, S.H., 2006. Does investor misvaluation drive the takeover market?


Doukas, J., 1995. Overinvestment, Tobin's q and gains from foreign acquisitions. Journal of Banking & Finance

19, 1285-1303.

Doukas, J., Travlos, N.G., 1988. The effect of corporate multinationalism on shareholders’ wealth: Evidence from

international acquisitions. Journal of Finance 43, 1161-1175.

Doukas, J.A., Petmezas, D., 2007. Acquisitions, overconfident managers and self‐attribution bias. European

Financial Management 13, 531-577.

Draper, P., Paudyal, K., 2006. Acquisitions: private versus public. European Financial Management 12, 57-80.

Draper, P., Paudyal, K., 2008. Information asymmetry and bidders’ gains. Journal of Business Finance &

Accounting 35, 376-405.

Eckbo, B.E., 2009. Bidding strategies and takeover premiums: A review. Journal of Corporate Finance 15, 149-

178.

Eckbo, B.E., 2014. Corporate takeovers and economic efficiency. Annual Review of Financial Economics 6, 51-

74.

Eckbo, B.E., Langohr, H., 1989. Information disclosure, method of payment, and takeover premiums: Public and

private tender offers in France. Journal of Financial Economics 24, 363-403.

Eckbo, B.E., Makaew, T., Thorburn, K.S., 2016. Are stock-financed takeovers opportunistic? SSRN Working

Paper.

Eckbo, B.E., Thorburn, K.S., 2000. Gains to bidder firms revisited: domestic and foreign acquisitions in Canada.


References

308

Edmans, A., Goldstein, I., Jiang, W., 2012. The real effects of financial markets: The impact of prices on takeovers.


Edmister, R., Walkling, R., 1985. Determinants of tender offer premiums. Financial Analysts Journal Jan.-Feb.,

27-37.

Edwards, W., 1968. Conservatism in Human Information Processing. In B. Kleinmuntz (Ed.), Formal

Representation of Human Judgment (pp. 17-52). New York: Wiley.

Erel, I., Liao, R.C., Weisbach, M.S., 2012. Determinants of cross‐border mergers and acquisitions. Journal of

Finance 67, 1045-1082.

Faccio, M., Masulis, R.W., 2005. The choice of payment method in European mergers and acquisitions. Journal

of Finance 60, 1345-1388.

Falaato, A., 2008. Superstars or superlemons? Top executive pay and corporate acquisitions, American Law &

Economics Association Annual Meetings: bepress, p. 102.

Fama, E., Fisher, L., Jensen, M., Roll, R., 1969. The adjustment of stock prices to new information. International

Economic Review 10.

Fama, E.F., 1998. Market efficiency, long-term returns, and behavioral finance. Journal of Financial Economics

49, 283-306.

Fama, E.F., French, K.R., 1993. Common risk factors in the returns on stocks and bonds. Journal of Financial

Economics 33, 3-56.

Fama, E.F., French, K.R., 1996. Multifactor explanations of asset pricing anomalies. Journal of Finance 51, 55-

84.

Fama, E.F., French, K.R., 2015. A five-factor asset pricing model. Journal of Financial Economics 116, 1-22.

Ferreira, M.A., Laux, P.A., 2007. Corporate governance, idiosyncratic risk, and information flow. Journal of

Finance 62, 951-989.

Fishman, M.J., 1989. Preemptive bidding and the role of the medium of exchange in acquisitions. Journal of

Finance 44, 41-57.

France, C.J., 1902. The gambling impulse. American Journal of Psychology 13, 364-407.

References

309

Franks, J., Harris, R., Titman, S., 1991. The post-merger share-price performance of acquiring firms. Journal of

Financial Economics 29, 81-96.

Franks, J.R., Broyles, J.E., Hecht, M.J., 1977. An industry study of the profitability of mergers in the United

Kingdom. Journal of Finance 32, 1513-1525.

Franks, J.R., Harris, R.S., 1989. Shareholder wealth effects of corporate takeovers: the U.K. experience 1955–

1985. Journal of Financial Economics 23, 225-249.

Franks, J.R., Harris, R.S., Mayer, C., 1988. Means of payment in takeovers: Results for the United Kingdom and

the United States, Corporate takeovers: Causes and consequences: University of Chicago Press, pp. 221-264.

Fu, F., Lin, L., & Officer, M. S., 2013. Acquisitions driven by stock overvaluation: Are they good deals? Journal

of Financial Economics, 109, 24-39.

Fuller, K., Netter, J., Stegemoller, M., 2002. What do returns to acquiring firms tell us? Evidence from firms that

make many acquisitions. Journal of Finance 57, 1763-1793.

Galasso, A., Simcoe, T.S., 2011. CEO overconfidence and innovation. Management Science 57, 1469-1484.

Gamble, K.J. and Johnson, B., 2014. How Prior Outcomes Affect Individual Investors’ Subsequent Risk Taking.

SSRN Working Paper.

George, T.J., Hwang, C.Y., 2004. The 52‐week high and momentum investing. Journal of Finance 59, 2145-2176.

Gervais, S., Heaton, J.B., Odean, T., 2011. Overconfidence, compensation contracts, and capital budgeting.


Ghosh, A., 2001. Does operating performance really improve following corporate acquisitions? Journal of

Corporate Finance 7, 151-178.

Golubov, A., Petmezas, D., Travlos, N.G., 2012. When it pays to pay your investment banker: New evidence on

the role of financial advisors in M&As. Journal of Finance 67, 271-311.

Golubov, A., Petmezas, D., Travlos, N.G., 2015. Do stock-financed acquisitions destroy value? New methods and

evidence. Review of Finance, rfv009.

Gorton, G., Kahl, M., Rosen, R.J., 2009. Eat or be eaten: A theory of mergers and firm size. Journal of Finance

64, 1291-1344.

References

310

Gould, W., 1993. Linear splines and piecewise linear functions. Stata Technical Bulletin 15, 13-17.

Graham, B., 2005. Intelligent Investor: The Classic Text on Value Investing. Harper Collins. (Original work

published in 1949).

Graham, J.R., Harvey, C.R., Puri, M., 2013. Managerial attitudes and corporate actions. Journal of Financial

Economics 109, 103-121.

Gregory, A., 1997. An examination of the long run performance of U.K. acquiring firms. Journal of Business

Finance & Accounting 24, 971-1002.

Gregory, A., 2005. The long run abnormal performance of U.K. acquirers and the free cash flow hypothesis.

Journal of Business Finance & Accounting 32, 777-814.

Gregory, A., O'Donohoe, S., 2014. Do cross border and domestic acquisitions differ? Evidence from the

acquisition of U.K. targets. International Review of Financial Analysis 31, 61-69.

Grinblatt, M., Keloharju, M., 2001. What makes investors trade? Journal of Finance 56, 589-616.

Grinstein, Y., Hribar, P., 2004. CEO compensation and incentives: Evidence from M&A bonuses. Journal of


Guay, W.R., 1999. The sensitivity of CEO wealth to equity risk: An analysis of the magnitude and determinants.


Guest, P.M., 2009. The impact of mergers and acquisitions on executive pay in the United Kingdom. Economica

76, 149-175.

Gugler, K., Mueller, D.C., Yurtoglu, B.B., Zulehner, C., 2003. The effects of mergers: an international comparison.

International Journal of Industrial Organization 21, 625-653.

Guo, J.M., Petmezas, D., 2012. What are the causes and effects of M&As? The U.K. evidence. Multinational

Finance Journal 16, 1-27.

Hagendorff, J., Collins, M., Keasey, K., 2008. Investor protection and the value effects of bank merger

announcements in Europe and the US. Journal of Banking & Finance 32, 1333-1348.

Hansen, R.G., 1987. A theory for the choice of exchange medium in mergers and acquisitions. Journal of Business

60, 75-95.

References

311

Harford, J., 1999. Corporate cash reserves and acquisitions. Journal of Finance 54, 1969-1997.

Harford, J., Humphery-Jenner, M., Powell, R., 2012. The sources of value destruction in acquisitions by

entrenched managers. Journal of Financial Economics 106, 247-261.

Harford, J., Klasa, S., Walcott, N., 2009. Do firms have leverage targets? Evidence from acquisitions. Journal of


Harford, J., Li, K., 2007. Decoupling CEO wealth and firm performance: The case of acquiring CEOs. Journal of

Finance 62, 917-949.

Hartzell, J.C., Ofek, E., Yermack, D., 2004. What’s in it for me? CEOs whose firms are acquired. Review of

Financial Studies 17, 37-61.

Hayward, M.L., Hambrick, D.C., 1997. Explaining the premiums paid for large acquisitions: Evidence of CEO

hubris. Administrative Science Quarterly, 103-127.

Healy, P.M., Palepu, K.G., Ruback, R.S., 1992. Does corporate performance improve after mergers? Journal of


Heath, C., Huddart, S., Lang, M., 1999. Psychological factors and stock option exercise. Quarterly Journal of

Economics, 601-627.

Heider, F., 1958. The psychology of interpersonal relations. New York: Wiley. p. 322.

Higson, C., Elliott, J., 1998. Post-takeover returns: The U.K. evidence. Journal of Empirical Finance 5, 27-46.

Hirshleifer, D., Low, A., Teoh, S.H., 2012. Are overconfident CEOs better innovators? Journal of Finance 67,

1457-1498.

Hoberg, G., Phillips, G., 2010. Product market synergies and competition in mergers and acquisitions: A text-

based analysis. Review of Financial Studies 23, 3773-3811.

Hodgkinson, L., Partington, G.H., 2008. The motivation for takeovers in the U.K. Journal of Business Finance &

Accounting 35, 102-126.

Houston, J.F., James, C.M., Ryngaert, M.D., 2001. Where do merger gains come from? Bank mergers from the

perspective of insiders and outsiders. Journal of Financial Economics 60, 285-331.

References

312

Huang, Y.-S., Walkling, R.A., 1987. Target abnormal returns associated with acquisition announcements:

Payment, acquisition form, and managerial resistance. Journal of Financial Economics 19, 329-349.

Huddart, S., Lang, M., Yetman, M.H., 2009. Volume and price patterns around a stock's 52-week highs and lows:

Theory and evidence. Management Science 55, 16-31.

Ismail, A., Krause, A., 2010. Determinants of the method of payment in mergers and acquisitions. Quarterly

Review of Economics and Finance 50, 471-484.

Jarrell, G.A., Poulsen, A.B., 1989. The returns to acquiring firms in tender offers: Evidence from three decades.

Financial Management, 12-19.

Jemison, D.B., Sitkin, S.B., 1986. Corporate acquisitions: A process perspective. Academy of Management Review

11, 145-163.

Jensen, M., 1986. Agency cost of free cash flow, corporate finance, and takeovers. Corporate Finance, and

Takeovers. American Economic Review 76.

Jensen, M.C., 1988. Takeovers: Their causes and consequences. Journal of Economic Perspectives 2, 21-48.

Jensen, M.C., 2005. Agency costs of overvalued equity. Financial Management 34, 5-19.

Jensen, M.C., Meckling, W.H., 1976. Theory of the firm: Managerial behavior, agency costs and ownership

structure. Journal of Financial Economics 3, 305-360.

Kahneman, D., 2003. Maps of bounded rationality: Psychology for behavioral economics. American Economic

Review 93, 1449-1475.

Kahneman, D., Tversky, A., 1979. Prospect theory: An analysis of decision under risk. Econometrica: Journal of

the Econometric Society 47, 263-292.

Kau, J.B., Linck, J.S., Rubin, P.H., 2008. Do managers listen to the market? Journal of Corporate Finance 14,

347-362.

Kim, J.-Y.J., Haleblian, J.J., Finkelstein, S., 2011. When firms are desperate to grow via acquisition: The effect

of growth patterns and acquisition experience on acquisition premiums. Administrative Science Quarterly 56, 26-

60.

King, D.R., Dalton, D.R., Daily, C.M., Covin, J.G., 2004. Meta‐analyses of post‐acquisition performance:

Indications of unidentified moderators. Strategic Management Journal 25, 187-200.

References

313

Kliger, D., Kudryavtsev, A., 2008. Reference point formation by market investors. Journal of Banking & Finance

32, 1782-1794.

Kohers, N., Kohers, G., Kohers, T., 2007. Glamour, value, and the form of takeover. Journal of Economics and

Business 59, 74-87.

Kuipers, D.R., Miller, D.P., Patel, A., 2009. The legal environment and corporate valuation: Evidence from cross-

border takeovers. International Review of Economics & Finance 18, 552-567.

Kumar, A., 2009. Who gambles in the stock market? Journal of Finance 64, 1889-1933.

Kumar, M., Dixit, J., Francis, B., 2015. The impact of prior stock market reactions on risk taking in acquisitions.

Strategic Management Journal 36, 2111-2121.

Lang, L.H.P., Stulz, R.M., Walkling, R.A., 1989. Managerial performance, Tobin's Q, and the gains from

successful tender offers. Journal of Financial Economics 24, 137-154.

Larcker, D.F., Rusticus, T.O., 2010. On the use of instrumental variables in accounting research. Journal of

Accounting and Economics 49, 186-205.

Lehn, K.M., Zhao, M., 2006. CEO turnover after acquisitions: Are bad bidders fired? Journal of Finance 61,

1759-1811.

Levis, M., 2011. The performance of private equity‐backed IPOs. Financial Management 40, 253-277.

Li, J., Yu, J., 2012. Investor attention, psychological anchors, and stock return predictability. Journal of Financial

Economics 104, 401-419.

Lin, H.-C., Chou, T.-K., Cheng, J.-C., 2011. Does market misvaluation drive post-acquisition underperformance

in stock deals? International Review of Economics & Finance 20, 690-706.

Lin, L., Lin, D., 2014. Agency costs of free cash flow and bidders’ long-run takeover performance. Universal

Journal of Accounting and Finance 2, 1-8.

Linn, S.C., Switzer, J.A., 2001. Are cash acquisitions associated with better postcombination operating

performance than stock acquisitions? Journal of Banking & Finance 25, 1113-1138.

Loderer, C., Martin, K., 1990. Corporate acquisitions by listed firms: The experience of a comprehensive sample.

Financial Management, 17-33.

References

314

Loderer, C., Martin, K., 1992. Post-acquisition performance of acquiring firms. Financial Management, 69-79.

Loughran, T., Ritter, J.R., 2002. Why don’t issuers get upset about leaving money on the table in IPOs? Review

of Financial Studies 15, 413-444.

Loughran, T., Vijh, A.M., 1997. Do long‐term shareholders benefit from corporate acquisitions? Journal of

Finance 52, 1765-1790.

Luo, Y., 2005. Do insiders learn from outsiders? Evidence from mergers and acquisitions. Journal of Finance 60,

1951-1982.

Lyon, J.D., Barber, B.M., Tsai, C.L., 1999. Improved methods for tests of long‐run abnormal stock returns.


Malmendier, U., Tate, G., 2005. CEO overconfidence and corporate investment. Journal of Finance 60, 2661-

2700.

Malmendier, U., Tate, G., 2008. Who makes acquisitions? CEO overconfidence and the market’s reaction. Journal


March, J.G., Shapira, Z., 1987. Managerial perspectives on risk and risk taking. Management Science 33, 1404-

1418.

Martin, K.J., 1996. The method of payment in corporate acquisitions, investment opportunities, and management

ownership. Journal of Finance 51, 1227-1246.

Martin, K.J., McConnell, J.J., 1991. Corporate performance, corporate takeovers, and management turnover.


Martynova, M., Oosting, S., Renneboog, L., 2006. The long-term operating performance of European mergers

and acquisitions.

Martynova, M., Renneboog, L., 2006. Mergers and acquisitions in Europe. SSRN Working Paper.

Martynova, M., Renneboog, L., 2008. A century of corporate takeovers: What have we learned and where do we

stand? Journal of Banking & Finance 32, 2148-2177.

Martynova, M., Renneboog, L., 2009. What determines the financing decision in corporate takeovers: Cost of

capital, agency problems, or the means of payment? Journal of Corporate Finance 15, 290-315.

References

315

Martynova, M., Renneboog, L., 2011. The performance of the European market for corporate control: Evidence

from the fifth takeover wave. European Financial Management 17, 208-259.

Massa, M., Zhang, L., 2009. Cosmetic mergers: The effect of style investing on the market for corporate control.


Masulis, R.W., Simsir, S.A., 2015. Deal initiation in mergers and acquisitions. SSRN Working Paper.

Mehra, R., Prescott, E.C., 1985. The equity premium: A puzzle. Journal of Monetary Economics 15, 145-161.

Mitchell, M.L., Lehn, K., 1990. Do bad bidders become good targets? Journal of Political Economy 98, 372-398.

Moeller, S., 2015. Why taking risks in M&A should pay off.

Available at https://blogs.intralinks.com/2015/07/why-taking-risks-in-mergers-acquisitions-should-pay-off/.

Moeller, S.B., Schlingemann, F.P., 2005. Global diversification and bidder gains: A comparison between cross-

border and domestic acquisitions. Journal of Banking & Finance 29, 533-564.

Moeller, S.B., Schlingemann, F.P., Stulz, R.M., 2004. Firm size and the gains from acquisitions. Journal of


Moeller, S.B., Schlingemann, F.P., Stulz, R.M., 2007. How do diversity of opinion and information asymmetry

affect acquirer returns? Review of Financial Studies 20, 2047-2078.

Morck, R., Shleifer, A., Vishny, R.W., 1988. Management ownership and market valuation: An empirical analysis.


Morck, R., Shleifer, A., Vishny, R.W., 1990. Do managerial objectives drive bad acquisitions? Journal of Finance

45, 31-48.

Morrow, J., Sirmon, D.G., Hitt, M.A., Holcomb, T.R., 2007. Creating value in the face of declining performance:

Firm strategies and organizational recovery. Strategic Management Journal 28, 271-283.

Mueller, D.C., 1980. The determinants and effects of mergers: An international comparison. Oelgeschlager, Gunn

& Hain; Königstein/Ts.: Verlag A. Hain.

Myers, S.C., Majluf, N.S., 1984. Corporate financing and investment decisions when firms have information that

investors do not have. Journal of Financial Economics 13, 187-221.

References

316

Northcraft, G.B., Neale, M.A., 1987. Experts, amateurs, and real estate: An anchoring-and-adjustment perspective

on property pricing decisions. Organizational Behavior and Human Decision Processes 39, 84-97.

Odean, T., 1998. Are investors reluctant to realize their losses? Journal of Finance 53, 1775-1798.

Olsen, R.A., 1997. Prospect theory as an explanation of risky choice by professional investors: Some evidence.

Review of Financial Economics 6, 225-232.

Palepu, K.G., 1986. Predicting takeover targets: A methodological and empirical analysis. Journal of Accounting

and Economics 8, 3-35.

Pachare S., 2010, Market Power, Extrapolation and the Cross-Section of Stock Returns. Draft.

Pettengill, G.N., Jordan, B.D., 1990. The overreaction hypothesis, firm size, and stock market seasonality. Journal

of Portfolio Management 16, 60-64.

Powell, R.G., Stark, A.W., 2005. Does operating performance increase post-takeover for U.K. takeovers? A

comparison of performance measures and benchmarks. Journal of Corporate Finance 11, 293-317.

Raj, M., Forsyth, M., 2003. Hubris amongst U.K. bidders and losses to shareholders. International Journal of

Business 8.

Rau, P., Vermaelen, T., 1998. Glamour, value and the post-acquisition performance of acquiring firms. Journal


Rhodes-Kropf, M., Robinson, D.T., 2008. The market for mergers and the boundaries of the firm. Journal of

Finance 63, 1169-1211.

Rhodes-Kropf, M., Robinson, D.T., Viswanathan, S., 2005. Valuation waves and merger activity: The empirical

evidence. Journal of Financial Economics 77, 561-603.

Rhodes-Kropf, M., Viswanathan, S., 2004. Market valuation and merger waves. Journal of Finance 59, 2685-

2718.

Roll, R., 1986. The hubris hypothesis of corporate takeovers. Journal of Business 59, 197-216.

Rossi, S., Volpin, P.F., 2004. Cross-country determinants of mergers and acquisitions. Journal of Financial


References

317

Ruback, R.S., 1988. Do target shareholders lose in unsuccessful control contests? Corporate takeovers: Causes

and consequences: University of Chicago Press, pp. 137-156.

Sacheti, A., Gregory-Smith, I., Paton, D., 2016. Managerial decision making under uncertainty: The case of

twenty 20 cricket. Journal of Sports Economics 17, 44-63.

Sanders, W.G., Hambrick, D.C., 2007. Swinging for the fences: The effects of CEO stock options on company

risk taking and performance. Academy of Management Journal 50, 1055-1078.

Savor, P.G., Lu, Q., 2009. Do stock mergers create value for acquirers? Journal of Finance 64, 1061-1097.

Schneider, C., Spalt, O.G., 2015. Acquisitions as lotteries? The selection of target-firm risk and its impact on

merger outcomes (January 9, 2015). Critical Finance Review (Forthcoming).

Schwert, G.W., 1996. Markup pricing in mergers and acquisitions. Journal of Financial Economics 41, 153-192.

Schwert, G.W., 2000. Hostility in takeovers: In the eyes of the beholder? Journal of Finance 55, 2599-2640.

Servaes, H., 1991. Tobin's Q and the gains from takeovers. Journal of Finance 46, 409-419.

Seth, A., Song, K.P., Pettit, R., 2000. Synergy, managerialism or hubris? An empirical examination of motives

for foreign acquisitions of US firms. Journal of International Business Studies 31, 387-405.

Seth, A., Song, K.P., Pettit, R.R., 2002. Value creation and destruction in cross‐border acquisitions: an empirical

analysis of foreign acquisitions of US firms. Strategic Management Journal 23, 921-940.

Shefrin, H., Statman, M., 1985. The disposition to sell winners too early and ride losers too long: Theory and

evidence. Journal of Finance 40, 777-790.

Shleifer, A., Vishny, R.W., 1988. Value maximization and the acquisition process. Journal of Economic

Perspectives 2, 7-20.

Shleifer, A., Vishny, R.W., 2003. Stock market driven acquisitions. Journal of Financial Economics 70, 295-311.

Smith, A., 1759. The Theory of Moral Sentiment (London: A. Millar).

Soegiharto, S., 2009. What drives the payment of higher merger premiums? Gadjah Mada International Journal

of Business 11, 191-228.

References

318

Stock, J.H., Wright, J.H., Yogo, M., 2002. A survey of weak instruments and weak identification in generalized

method of moments. Journal of Business & Economic Statistics 20, 518-529.

Stulz, R., 1988. Managerial control of voting rights: Financing policies and the market for corporate control.


Sudarsanam, S., Mahate, A.A., 2003. Glamour acquirers, method of payment and post‐acquisition performance:

The U.K. evidence. Journal of Business Finance & Accounting 30, 299-342.

Switzer, J.A., 1996. Evidence on real gains in corporate acquisitions. Journal of Economics and Business 48, 443-

460.

Trautwein, F., 1990. Merger motives and merger prescriptions. Strategic Management Journal 11, 283-295.

Travlos, N.G., 1987. Corporate takeover bids, methods of payment, and bidding firms’ stock returns. Journal of

Finance 42, 943-963.

Tversky, A., Kahneman, D., 1974. Judgment under uncertainty: Heuristics and biases. Science 185, 1124-1131.

Tversky, A., Kahneman, D., 1992. Advances in prospect theory: Cumulative representation of uncertainty.

Journal of Risk and Uncertainty 5, 297-323.

Uysal, V.B., 2011. Deviation from the target capital structure and acquisition choices. Journal of Financial

Economics 102, 602-620.

Uysal, V.B., Kedia, S., Panchapagesan, V., 2008. Geography and acquirer returns. Journal of Financial

Intermediation 17, 256-275.

Vermaelen, T., Xu, M., 2014. Acquisition finance and market timing. Journal of Corporate Finance 25, 73-91.

Vijh, A.M., Yang, K., 2013. Are small firms less vulnerable to overpriced stock offers? Journal of Financial


Wansley, J.W., Lane, W.R., Yang, H.C., 1987. Gains to bidder firms in cash and securities transactions. Financial

Review 22, 403-414.

Weitzel, U., Kling, G., 2014. Sold below value? Why some takeover offers have negative premiums. SSRN

Working Paper.

References

319

Wulf, J., 2004. Do CEOs in mergers trade power for premium? Evidence from “mergers of equals”. Journal of

Law, Economics, and Organization 20, 60-101.

Zhang, L., 2009. Why do firms pay cash in acquisitions? Evidence from a demand perspective. SSRN Working

Paper.

Zhang, P.M., 2001. What really determines the payment methods in M&A deals. Manchester School of

Management, Working Paper.

the reference point effect, m&a misvaluations and merger...

Documents